QSAS

[BoStros]

I created this earlier today. Hopefully it can lead to the end of the W-L record having any significance. Looking for thoughts, opinions and ideas about how to make this stat a better complex and accurate tool.

 

 

Quality Start Applicable To Season

 

 

My own version of the "quality start" stat.

 

I basically took the old QS and changed the ER requirements so that a pitcher must pitch below the league average on earned runs allowed pending on the amount of innings played in order to qualify. You gotta be one step ahead of the MLB opposition. Read further for deeper analysis.

 

A very simple formula actually. Just fills up the hole left by John Lowe.

 

In order to get a league average, you must first divide the amount of innings needed into the amount of innings pitched within the entire MLB in that season. After you have that quotient, divide it into the full amount of ER given up that entire season, estimate, and you have your league average. Now, for an example....

 

We will find the league average for every 7 innings pitched....

 

2009 MLB IP - 43,272

2009 MLB ER - 20,779

 

 

43272/7 = 6182

 

20779/6182 = (estimated to nearest whole) 3

 

Now what this means is that the average pitcher would give up 3 earned runs every 7 innings pitched. Now if you wanna win the game, you gotta play above the opposition meaning that in order to get a QSAS if you pitched a 7 inning start in 2009, you must give up 2 runs or less.

 

Here's a 2009 list of requirements -

 

6 IP - 2 runs or less

7 IP - 2 runs or less

8 IP - 3 runs or less

9 IP - 3 runs or less

 

 

 

Criticize me, support me whatever you feel needs to be said, please, say it.

 

pitcherevaluation.blogspot.com

[Dojji]

With the huge proportional difference between the 3 and 2 when you're talking about a percentage or ratio spread over multiple samples, I figure that you might want to be a bit careful about arbitrarily forcing the standard down another run. A season full of traditional quality starts would net a starter a 4.50 ERA. Your stat holds them up to a Cy Young caliber standard. May want to think about that -- unless you think that the "quality start" stat is too inclusive anyway.

[Dipre]

The idea is solid, but the calculation is fundamentally wrong, because you're not separating the source of the IP for your formula. Starting pitchers and relievers are completely different beasts, so what you'd have to separate both IP and ER from starters, which is entirely possible by using the following Fangraphs link:

 

http://www.fangraphs.com/teams.aspx?pos=all&stats=sta&lg=all&type=0&season=2009&month=0

 

Export to Excel, and calculate from there.

 

Using this separation of starters and relievers, you come up with the following results:

 

2009 Starter's IP: 28,251

 

2009 Starter's ER: 13,977

 

First, ERA: 2009 Starter's ERA: 4.45

 

Now let's use your formula to determine how many runs could a pitcher give up from innings six to nice for a start to be considered a "Quality start".

 

28,251/6: 4,708.5

 

28,251/7: 4,305.8

 

28,251/8: 3531.3

 

28,251/9: 3139

 

 

6 IP: 13,977/4708.5: 2.96 (Round up to 3)

 

7 IP: 13,977/4305.8: 3.24 (Round down to 3)

 

8 IP: 13,977/3531.3: 3.95 (Round up to 4)

 

9 IP: 13,977/3139: 4.45 (Round down to 4)

 

It's essentially the same result as the typical quality start, but for pitching eight or nine innings, allowing four runs instead of three would still result in a quality start, all of this utilizing the League average as a benchmark. Even more interesting is the fact that the league average ERA is exactly the 4.45 mark that you would receive from an average pitcher's nine-inning start.

[Keeper]

I've always wondered why 4 runs in 8 or 9 innings isn't considered a quality start under the current criteria.