What Happened to the Quality Start & Why It Shouldn’t Disappear

In the current war between traditionalists and sabrmetricians, there have been a few casualties. Families have been driven apart by the usefulness of OBP over AVG and severe insults are hurled every day on twitter against and in defense of WAR. One of the casualties rarely noticed has been the quality start.

The Quality Start as a stat was developed in 1985 by John Lowe and simply notes that a pitcher went six or more innings and allowed three or less runs. While there are issues with this, including pitchers who throw seven or more innings and allow four or less runs and those who pitch poorly early on, but somehow manage to keep the other team, it still answers the basic question of “did the pitcher do his job?”

Most stats in baseball are are the cumulative results of single events. Counting stats like home runs and RBI and average stats, like AVG/OBP/SLG are all of this kind. Even WAR, BABIP, WRC+ and FIP are just equations that look back at events that occur one at bat at a time. The type of stat that answers the question named above are much more rare and are falling out of favor. A save answers the question “did the closer do his job?” If the closer’s job is simply to record three outs in the ninth inning with a lead of three or less runs, then saves and save percent is a good indicator of whether he did his job. The hold is the same way. Even more so than ERA (which can be increased to an extreme level from one blown save or non save situation) or FIP (which is a better predictor of whether or not they will be successful in the future) these stats describe whether a particular player has come through when they were needed.

For the starting pitcher, while not perfect, the quality start was the best judge of whether or not they did their job. While die hard old schoolers prefer to use the pitcher win as this marker, that stat assumes that pitchers have some control over the offense performance of his team, even in the American League where he doesn’t bat. The quality start instead assumes that a Major League Baseball team should be able to score at least four runs in a game. This number is arbitrary, but in innings and runs, but at least it gives us a starting point.

Above is a partial list of the American Leaders in quality starts (from ESPN since MLB.com no longer features the stat on it’s leaderboards) and there is no question that these are among the best pitchers in the league. Important to note is that there is little correlation between the three numbers above. Just because a player pitches six innings and allows less than three runs often doesn’t mean his team will support him (both Jose Quintana and Corey Kluber have half or less the wins compared to QS) or that he will do it every time out.

The most perfect example of this dichotomy was Ubaldo Jimenez for the Indians in 2012. While his end of the season numbers were absolutely terrible (9-17, 5.40 ERA) and reflective of his overall performance, he still made 13 quality starts that season (similar to what Danny Salazar has done with a 3.30 ERA in 2015). His traditional stats didn’t take into account the fact that when he was good, he was great and when he was bad, he was terrible. In that season there were three games where he left the game after the sixth with no runs scored and nine with two or less. No one would expect a pitcher with a 5.40 ERA to give his team a situation that should be nearly an automatic win (with a good bullpen) more than ten times a year, but Jimenez did.

This year, the Indians lack of run support in general (and for Kluber in particular) has lead to some very unjust win-loss totals. That the 14th best pitcher in ERA in the AL is 31st in wins is enough to ignore that stat completely. We shouldn’t, however, look past the quality starts. For the sake of the Indians poor offense, we can even look beyond quality starts to games where a pitcher has thrown at least six innings and allowed two or fewer runs.

For further explanation the stats tallied are total starts, starts with at least six innings and two or less earned runs, then starts with at least six innings and one or fewer earned runs. In modern baseball, a start of six innings with one or less run allowed should be as close to an automatic win as is possible and that has generally been the case with Kluber earning a win in 6 of 8, Carlos Carrasco 6 of 8, Salazar 9 of 10 and Trevor Bauer 4 of 8. While no pitcher has been handed a loss in this particular situation, they have in their less stellar performances that were still good enough for a Quality Start.

One of the primary purposes for this whole deliberation is to show the greatness of Bauer in particular.  While he has, by far, the worst numbers among the regular four considering ERA, strike outs, FIP or essentially any other metric, he has been as good at keeping the Indians in games as any pitcher. The Indians have eventually lost 13 games that Kluber has started and pitched at least six innings while allowing four or less runs. They have only lost 14 of Bauer’s starts in total despite the fact that eight times he has allowed five or more runs, only in one case pitching beyond the fifth.

This is not trying to insinuate anything, but instead adds another piece to the puzzle. For a team with limited offense, a pitcher with an ERA above 5.00 who allows three runs in six innings nearly every single time out is worth less than a pitcher who allows less than two runs through six regularly, but gets blown up every third or fourth start. At least he gives the opportunity to win those in between.

This is the value of the quality start. It shouldn’t decide the Cy Young, it can’t be used to predict the future or as the sole judge of any pitcher, but it can be used to qualify other numbers, like ERA (and on a lesser level, wins and losses). Before attacking Bauer for his poor season numbers, at least give him credit for giving the Indians a great chance to win just as often as every other pitcher on the staff.