14 replies to this topic

[Brett]

I was thinking about the quality start cut-off (3 or fewer earned runs in 6 or more innings), as I am a big fan of the quality start stat.

What I got thinking about is whether a guy who gives you the bare minimum for a quality start is better or worse than a guy who alternates dominant starts and terrible ones.

So, for example, which pitcher would you rather have:

Pitcher A has 30 starts, 180 innings, has given up 90 earned runs, and, thus, has a 4.50 ERA. In every single start, he's gone six innings, giving up exactly three runs in each start. Thus, he's a perfect 30 for 30 in quality starts.

Pitcher B also has 30 starts, 180 innings, has given up 90 earned runs, and, thus, also has a 4.50 ERA. In 15 of his starts, he's thrown a shutout - 9 innings, 0 earned runs (which guarantees that his team won, except maybe once or twice when the team lost in extras (thanks, Carlos)). In the other 15 starts, he went just three innings, giving up six earned runs in each start. He's just 15 of 30 in quality starts, but his good starts were dominant, and his bad starts were terrible.

So, who are you picking?

[Fishin Phil]

My gut says pitcher B, but my head says pitcher A. On those terrible starts that pitcher B has, you chew up your bullpen. I guess it comes down to what type of offense you have to go with the pitcher. If your team is consistently putting up 4 or 5 runs per game, pitcher A would look great. If your team is struggling to scratch out 2 or 3 runs per game, you have a better chance with pitcher B.

[Crockett]

You always take the guaranteed win. So pitcher B.

[gblan014]

Unless we have the ability to know in advance what day pitcher B will be having (9 inning shutout or 3 inning, 6 runs allowed); I'd go with pitcher A because of the predictability factor. You'd always know what to expect with pitcher A, theoretically, which helps you plan out your bullpen and your bench.

[Kansas Cubs Fan]

Whats the formula for ERA?

[TWC]

Whats the formula for ERA?

It's the measure of earned runs given up by a pitcher per 9 innings pitched.

Earned runs x 9 / IP.

Now the ERA+ calculation, that's gonna take a bit longer to calc...

[Dave]

A lot of factors come into play for me.

With Pitcher A, you need to make sure you have an offense that will score 4+ runs and a bullpen that will hold on for the last 3 innings. But that kind of predictability would be nice.

With Pitcher B, you'd need to have a deep bullpen, because having to get 6 innings out of it (plus whatever the rest of the rotation leaves it) is going to be a killer. But you're also getting 15 guaranteed wins a year.

If I had to make a choice, I'd go with B, likely a higher WAR, therefore (statistically) likely more valuable to the team.

[NormB]

For results, I'll take Pitcher A. At least you're in the game EVERY time.
Pitcher B is nearly a lock to put you in a hole within 3 innings.

[Tommy]

I'd take pitcher A. Pitcher B would just give me high blood pressure. If you could get a team full of players that are completely constant with their production, just think how easy that would be to manage!

[Cubbie Blues]

Haven't we already seen this as a closer? Does anybody remember Mitch Williams? Nobody ever knew if he would be throwing down the plate or the batters box any given day.

[Dave H]

I would look at consistency. If I was to know that "Pitcher A" was taking the be bump every 5 days, I'd feel we have a chance to win everytime. Pitcher B would have me look at the bullpen status after the first pitch.

[Ron]

Brett, are you sort of abstractly comparing Demptster (A) and Wells ( from last year? Now obviously it is not a perfect comparison but from generalities it is close. Dempster was Mr. Minimum quality start last year and it seemed that Wells either got completely shelled early or was great until Quade left him in an inning too long. I think almost everybody would pick Dempster. I have always liked Wells and don't think he should be scrapped just yet. He did have a great spring training last year and I don't think he was ever really healthy until the end.

[FFP]

It is January.
(I bet these "thought experiments" are what chess masters do while waiting in line at the grocery store instead of reading The Enquirer)
My first thought is: don't say anything bad about the Quality Start stat because Brett likes it. My second thought is: I'm going to find out why.

Now, third thought, I take pitcher B an pencil in my .500 win rate for his spot on the staff. There are so few iterations for a pitcher that I could come up on the short end on many more of pitcher A's games. It doesn't seem just. It just seems right.

[Brett]

Brett, are you sort of abstractly comparing Demptster (A) and Wells (B ) from last year? Now obviously it is not a perfect comparison but from generalities it is close. Dempster was Mr. Minimum quality start last year and it seemed that Wells either got completely shelled early or was great until Quade left him in an inning too long. I think almost everybody would pick Dempster. I have always liked Wells and don't think he should be scrapped just yet. He did have a great spring training last year and I don't think he was ever really healthy until the end.

It's interesting to put players' faces on it, but no, I hadn't been thinking of anyone in particular when I brought this up. Dempster is very interesting, though, because yeah - he was all about the 6 inning, 3 ER starts last year.

[SirCub]

From a pure math standpoint, looking at only the 30 games each player starts, the answer is clearly player B. Dugout Central compiled statistics from a 5-year period from baseball-reference.com, and show that teams that get the bare minimum quality start (6 IP, 3 ER) actually have a winning percentage of 48.3%, while a shutout (9 IP, 0 ER) is 98.4%, 3 IP with 6 (or more) runs is 17.5%.

So the expected wins for each player would be:

.483 WP X 30 G = 14.49 Wins for Player A

and

.984 X 15 G + .175 X 15 G = 17.3 Wins for Player B

Wins for player B are likely underestimated as well, because the .175 winning percentage was based on games where the pitcher pitched 3 innings and allowed 6 earned runs or more. So, based on this analysis, pitcher B is probably good for 3 more wins that season. However, to investigate this more thoroughly, the effects of these starts on the following games should be addressed. I think it would be interesting to look at the winning percentage of the teams in the games immediately following a start (say, the next 1-4 games?) in which their starting pitcher pitched 3, 6, or 9 innings, to see the effect of having to use your bullpen for 6 innings, 3, or not at all.