In yesterday’s Game Chatter we were discussing Quality Starts (QS), which began when someone asked if a game where the starter had 6 IP and 3 ER was still a quality one. My immediate answer was yes. But before we get into details on that, it should be pointed out the problem with the question.
Everyone always points to the bare minimums to qualify for a QS, which for most years is not a definition that we’d use for quality. But the simple fact is that the overwhelming majority of QS are not of the bare minimum. Case in point, the Mets have 26 QS this year and how many of those do you think were of the 6 IP, 3 ER variety? That’s right – four. So, we’re talking about 15% at the bottom of the necessary thresholds.
To date, the Mets are 21-5 when their SP gives a QS.
Some think that the run environment today means the QS needs to be revamped. But the average runs per game by an individual team in 2022 is 4.33 so far this season. A decade ago in 2012, that number was 4.32 and a score ago, in 2002 – in the middle of the Silly Ball era – there was a 4.62 rpg. The higher the run scoring environment, the tougher it is to throw a QS. This year’s run environment would have to be significantly worse than in the past to merit tinkering with the QS minimums. And they’re certainly not different enough from 2012 to merit any change.
With all of that said, let’s dive into a QS and what it means.
First, let’s start from history. Jump with me in the way back machine, where we travel to the first year of Mets360 when we looked into QS during the 2010 season. You can see the article here but the key takeaways were that the from 2007 to August 17 of 2010, “the Mets have 317 QS with a 176-50 record. They get a decision in 71 percent of the team’s QS and have a winning percentage of .779 when their starter goes at least 6 IP and gives up 3 ER or fewer.”
How do those baselines from more than a decade ago stand up to today? Here are how the Mets and their SP have fared in QS this season:
| Name | Quality Starts | QS Record |
|---|---|---|
| Chris Bassitt | 9 | 5-3 |
| Carlos Carrasco | 5 | 5-0 |
| Tylor Megill | 1 | 1-0 |
| David Peterson | 1 | 1-0 |
| Max Scherzer | 5 | 2-1 |
| Taijuan Walker | 5 | 5-0 |
| Totals | 26 | 19-4 |
Starters on the 2022 Mets have 26 QS with a 19-4 record. They get a decision in 88.5% of the team’s QS and have a winning percentage of .826 when their starter goes at least 6 IP and gives up 3 ER or fewer. So, the Mets are doing better across the board in QS than what they did when we first looked at the subject. However, the 2022 Mets are a significantly better team than their 2007-2010 counterparts. The best team in that span won 89 games and two of the four teams finished with sub-.500 records. The current team is on pace for 104 wins.
We would expect this year’s Mets team to be better in QS. The question is if the 26 ones the club has posted so far is an adequate sample size to compare to the 317-sample of the earlier period. Maybe, maybe not. Either way, it seems odd to me the idea that we should perhaps make it harder to get a QS when the results, specifically for the 2022 Mets, are already so good with the current definition.
In the past, people have suggested making the QS on a sliding scale, where the IP and ER requirements change depending upon how many innings you throw. If you go 6 IP then you need to give up only 2 ER while if you go 7 IP you can give up 3 ER and if you go 8 IP then you can give up 4 ER. Maybe if we employed a definition of this type, it would increase the winning percentage a bit. But, really, how much better does the winning percentage have to be from the .779 mark from the original sample? We’re looking for “quality,” not perfection.
To compare the QS to a newer metric, let’s look at barrels. To qualify for a barrel, a batted ball has to have a launch angle and exit value that produces at least a .500 AVG and a 1.500 SLG. Over at mlb.com in the glossary for barrels, it notes the following:
But similar to how Quality Starts have generally yielded a mean ERA much lower than the baseline of 4.50, the average Barrel has produced a batting mark and a slugging percentage significantly higher than .500 and 1.500, respectively. During the 2016 regular season, balls assigned the Barreled classification had a batting average of .822 and a 2.386 slugging percentage.
While some might note that barrels have a sliding scale incorporated into its definition, QS doesn’t need that to produce excellent results. In a way, it’s like using OPS. Mathematically, it’s a mistake to add things with different denominators. But we do it with OPS because A) it’s easy and B) it gets us 95% or more to the right answer.
Easy and accurate describes QS pretty well, too.
Another interesting point, at least in my opinion, regarding the QS metric is how it’s a bit anachronistic when compared to how many innings a typical starting pitcher goes today when compared to the past. Case in point, not a single team’s starting staff is averaging more than 5.6 innings per start so far this season: https://www.baseball-reference.com/leagues/majors/2022-starter-pitching.shtml
Does this fact actually add more value to what we consider a quality start?
The purpose of the QS is to look at games where a SP gave his team a chance to win. Since they lowered the mound in 1969, every year the average rpg is over 3.5 – so if you give up fewer runs than that, you’re giving your team a chance to win.
People get hung up on the bare minimums but those are a small subset of games. It’s like complaining about doubles because the one Pete Alonso got the other day was a bloop. The majority of doubles are hard-hit balls.
The Mets led MLB in innings pitched in 2019 and they averaged 5.8 innings per start so I’m not sure it’s significantly different in 2022.
I really don’t think a quality start giving up three runs in six innings is a good start. That is a 4.50 ERA. It used to be three runs in seven innings. Maybe we should go to two runs in five innings.
Per Fangraphs and my calculations, the average starter completes somewhere between 15 -16 outs per starts (5 -5 1/3 innings) in 2022. Six innings therefore is an above average start. The average starter has a 4.11 ERA. Three runs in six innings is a 4.50 ERA therefore that would be a below average start. So, 2 ER for 5 1/3 innings, or 3.38 ERA, would be the threshold for an above average start. That is average not quality. Quality is defined as a grade of excellence so therefore two run in six innings should be the definition.