The Quality Start (Criteria)

[BohKnowsBmore]

In tonight's game, John Smoltz lamented the criteria for the QS, specifically that it has a six inning requirement.  He stated that he'd like to see it moved up to seven innings so that it'll be a lot more rare.  I've heard this POV before (one of the guys on Baltimore radio always talks about it disdainfully). 

This misses the utility of the stat IMO.  Maybe people are bothered by the terminology, but it's basically a strong/decent start.  Having the requirements as not too stringent allows there to be a wider distribution of numbers. If it were 7IP and 2ER, for example, it would be Clayton Kershaw with a ton, maybe Dallas Keuchel, then a bunch of guys with really low numbers or zero.  The current iteration makes things more differentiated among pitchers, thus making the stat more useful IMO.

Interested to hear everyone else's thoughts on this.  It's not the biggest deal in the world, but I've heard this a fair amount and it seems to miss why the thresholds make sense.

[SteveA]

I agree with you.   I think it's a reasonable, rough yardstick of whether a starting pitcher gives his team a decent chance to win.

Here's how I see it:  take the absolute worst possible QS, 6 innings/3 runs.   A lot of QSs are better than this, of course.   But at 6 innings, 3 runs, the bullpen has to go 3 innings.   A half decent bullpen will have an ERA of 3.33 or better, so if your pitcher has a QS you probably will allow 4 runs in the game more often than not.   A half decent offense will average at least 4 runs per game.  So if you have a solid pen and at least average offense the worst QS gives you at least a 50% chance to win, and of course better QSs increase that chance.

Now say you fail to get a QS.   If you give up more than 3 runs, then unless your bullpen is perfect, you are going to give up 5+ runs in the game.   Unless you have a great offense, you don't have a 50% chance to win when you give up 5.

Or say you give up 3, but fail to go 6.   Now your bullpen has to pitch additional inning(s), and probably use guys who are not among the top 3 or 4 pitchers in the bullpen.  Unless you have one of the best bullpens in baseball, your runs-allowed expectaion for the game now goes over 4.   You probably no longer have a 50% chance to win unless you have a really good offense.

So to me the QS is an excellent rule of thumb indicating whether a pitcher has given his team a >50% chance to win, assuming they have at least a middle of the pack bullpen and middle of the pack offense.   Failing to get a QS knocks the percentage under 50% unless you have a great bullpen and/or offense.   I find that useful.

Obviously the stat doesn't take into account defense, ballpark, opposing lineup.   So it's not a perfect stat (I happen to believe there are no perfect stats, you just have to understand what each stat is trying to measure and take it for what it is).   That's why I called it a "rule of thumb" above rather than a stat.   It's a rough gauge of how often the starter is putting his team in position to win, and it takes into account both runs allowed AND duration, which are both important IMO.  

I guess the word "quality" really bugs a lot of people.   Maybe if another name has been chosen the stat wouldn't bother so many people.  Call them "Adequate Starts".   People who harp on "but it's a 4.50 ERA!" are kind of ignoring the fact that that is the absolute worst QS and that many are of course better.  You have to draw a line somewhere and I feel this is a good spot for the reasons I state above... any QS makes it fairly likely you will allow 4 runs or less; a failure to get a QS makes 5+ more likely, and in a run environment where the average team scores 4.something per game, that is a huge difference to me.

The line being drawn at a 4.50 ERA isn't nearly as bad as the save, where the worst save is basically an ERA of 18.00 (coming in with a 3 run lead in the 9th and giving up 2 runs).   (I realize that there are 3 inning saves where you could have a 20 run lead and give up 19, but in terms of evaluating 9th inning closers, the worst save has a 18.00 ERA).

[DirtyBird]

I don't care

[Frobby]

It's an arbitrary statistic, but good enough.    

[El Gordo]

I think QS% along with IP/G are a good quick measure of a SP''s effectiveness. The problem with the stat is semantics. They should call it an effective start or a good start.