Friday, April 30, 2010

An argument against the QS statistic

The purpose of this blog posting is to defend the position that the QS misrepresents the actual performance of a pitcher and that a different approach to measuring a pitcher's performance should be considered.

The problem with the QS is that there is a matter of inconsistency with regard to representing the actual performance of the pitcher's outing. For example, consider two pitchers

Pitcher A gives up 3 runs over 6 full innings and is pulled from the game. His final ERA is 4.50. He is awarded a QS.

Pitcher B gives up 4 runs over 9 full innings. His final ERA is 4.00. He is not awarded a QS.

One could easily argue that Pitcher B had a better game based on these measurements.

Consider also the situation,

Starting Pitcher A gives up 3 runs in the 1st the completes 6 full innings and is pulled from the game. Again, his ERA for that game is 4.50. He is awarded a QS.

Starting Pitcher C pitches 8 2/3 with 26Ks and 3 walks, then gives up a grandslam. He then Ks the last batter. His ERA is 4.00 with 27 Ks. He is not awarded a QS.

Starting Pitcher D pitches 5 no hit innings and is sent to the bench. Granted, this is unlikely but not out of the realm of possibility. Again, no QS is awarded.

Now clearly, over the 6 comparable innings, Pitcher C was exceptional and outperformed Pitcher A, yet was not awarded a QS.

Even Pitcher D outperformed Pitcher A and is not awarded a QS.

These scenarios illustrated that the QS statistic is not representative of the pitching performance in some instances.

Also, since decision to pull a pitcher is in the hands of the manager, the manager can actually have an effect on whether or not a pitcher is awarded a QS. This further supports the claim that the QS is not an accurate representation of the pitcher's performance. It is also representative of the manager's performance.

I would propose that the QS be awarded if a pitcher posts an ERA of 4.50 or less and must complete 5 innings. This normalizes the statistic to include the remaining innings. Of course, I can't change the MLB measuring system.


Now I agree that the 2009 system that allowed closers to be awarded wins was somewhat favorable for closing pitchers vs. starting pitchers not typically being awarded saves. I understand the need to ameliorate this situation.

Instead of the 2010 system what about keeping the W category and adding the L category to balance the closer getting the win. Closers usually have W/L records close to 1:1.

This approach would also cut back on the streaming pitcher approach that was, for some reason, determined to be unfavorable for this league. As pitchers in the FA pool are more likely to lose than pitchers on a stable roster.

I would like to hear some feedback on this as I'm sure I've overlooked something.

Tude Smacker Owner

Jesse Ward

9 comments:

Josh Coleman said...

The possibility of a pitcher completing a game and allowing 4 runs has been the basis for those detractors of the QA stat for years. I understand the aggervation of it but unless MLB decides to change the requirements little can be done to account for such an event. Yes management plays a big part in the chance at a possible QS. But what doesn't the MGR effect? MGR dictate lineups which factors into Run and RBI totals a MGR often dictates when a player has the green light to steal. The loss suggestion makes less since then just going back to the W. If we included the L stat then the 8 inning 1 run gem but no offensive support would count toward a negative stat. I am up for any suggestions regarding stats scoring but I really favor the QS over the W.

jesseward said...

Josh,

With all due respect, you are changing your story. The reason we went to the QS was to balance the closing pitcher situation. That situation was, a closing pitcher can get a win but a starter can't get a save. The W/L approach helps solve that problem and helps with the QS issue. Granted it doesn't solve either but it helps both.

Also, the manager certainly doesn't affect whether a player hits the ball or throws a strike or makes an error. The manager has an indirect effect on those parts of the game but has a direct effect on the QS situation because he can make the decision whether you preserve a QS or take a chance at not getting one. This is dissimilar to putting someone in the lineup and they get a bunch of hits.

So you don't want to get a pitcher that has problems getting run support?..... don't pick pitchers on shitty teams.

Jesse

Josh Coleman said...

How does the W/L scoring eliminate the Closer issue exactly. Your theory is that closers often have a similar W/L record. But how many good pitchers fall under the same scenario. A Win is a completely random stat and isn't as easy as just picking pitchers on good teams. Many 15 and 20 game winners have played for teams well below .500. The Hall of Fame is littered with pitchers who consitently played on poor teams. Also good pitchers who play on good teams don't always have the W's on the resume. The only thing you can do as a Fantasy Owner is pick pitchers who have low ERA and WHIP and high K totals. The QS gives a starting pitcher a stat that a closer cannot achieve. Just as the Save is exclusive to Closers.

jesseward said...

I'm not 100% sure what you are talking about.

I guess that means you win the argument.

jesseward said...

Ney, I've decided to address this piece by piece since that is what this forum is for, to discuss sports with valid reasoning and poor grammar. (I'm guilty of poor grammar too, hence why I stopped nitpicking. More importantly, nitpicking grammar is beside the point and therefore a waste of time)

1)"How does the W/L scoring eliminate the Closer issue exactly."
a) It doesn't and I never said it would/did. The W/L balances the advantages a closer has by getting a win. If he also receives losses, then he is essentially penalized for losing. Since most closers have ~1/1 W/L record, it balances out, so there is no advantage.

2) "But how many good pitchers fall under the same scenario."
a) We'll we would need to define "good pitcher" but of the top 40 with respect to wins, not a single pitcher has a 1/1 W/L ratio; all are better than .500. Granted a handful are ~.400 but these pitchers should simply be avoided if the L is a measurement for scoring.

3) "A Win is a completely random stat and isn't as easy as just picking pitchers on good teams."
a) You offer zero support that a win is a "completely random stat". In fact, if it were "completely random", all pitchers would have generally the same statistics regarding W/L ratio. Heck, I would have the same chance of pitching a win as would Greinke, if it were completely random. I think you'll agree it is indeed not completely random but in fact likely that the better the pitcher, the more likely it is that he will get the win.

4) "The Hall of Fame is littered with pitchers who consitently played on poor teams."
a) This is irrelevant to the point. The Hall of Fame is also littered with pitchers who played on great teams. Bottom line: If you want pitchers who get a bunch of wins, pick great pitchers on great teams to increase that likelihood.

5) "Also good pitchers who play on good teams don't always have the W's on the resume."
a) Granted they don't always, but frequently they do. Again, this point is irrelevant.

6) "The QS gives a starting pitcher a stat that a closer cannot achieve. Just as the Save is exclusive to Closers."
a) I am aware of that as I mentioned in my post.

Jess

jesseward said...
This comment has been removed by the author.
jesseward said...

Edit: A handful of the top 40 pitchers are .600 W/L.

~.400 L/W.

For clarification.

Anonymous said...

jesse,

you do have as many wins as greinke. keep up the good work.
i like the QS tho.

-todd

jesseward said...

Todd,

That's true but Greinke has a better chance of getting a win.

Why do you like the QS?

Jesse