MASTER NOTES: Master Notes notes noted in the Master Notes notebook

In a scattered week, let’s look at some Master Notes notes from the Master Notes notebook.

Fun With PQS: Wins Are The Dumbest Stat And Should Be Abolished in Fantasy Leagues, and While We’re At It, So Should Saves

And so should extremely long headings, if you ask me.

I love the Pure Quality Starts (PQS) logs at They’re part of the Leading Indicators suite ($) at, and they’re great for finding things to make my blood boil, when I need something because Twitter is down.

For instance, the PQS logs show wins are dumb. This is a position I've taken at some length in past Master Notes columns, like here, here, and here. I'll also take it up with you should be unwise enough to bring it up around the firepit at First Pitch Arizona.

To sum up those earlier diatribes, wins are a terrible measure of pitcher performance, because they too often reward the undeserving while punishing the excellent. They’re kind of like the Emmys that way.

The 2019 PQS logs once again say I’m right, which is one reason I like them. According to the PQS logs, 32% of all starting-pitcher starts have been wins, 35% no-decisions, and 34% losses. Projecting a category in this environment is little more than a coin toss, and doubly aggravating because it's so hard to find a coin with three sides.

In 12 games this season, starting pitchers:

  • Went fewer than six innings
  • Gave up six or more earned runs, one as many as eight
  • Gave up five or more hits, with three games allowing 11
  • Gave up as many as five walks
  • And had no more than eight Ks, including one with no Ks at all.

And you’ll probably guess the outcome. All those awful pitching performance resulted in wins! The booby prize went to Antonio Senzatela of COL, who just last month gave up eight earnies in 5 1/3 innings to get a richly undeserved win against CIN (and the fact that I’m a long-suffering Reds fan doesn’t factor into my chagrin).

In another 29 games, pitchers got wins while going fewer than six innings and giving up four-plus runs.

The common denominator in almost all of these travestial wins was an offensive outburst by the pitcher’s team—in Senzatela’s debacle, he gave up his eight runs while his teammates happened to be scoring 10. Benefiting handsomely from the work and success of others has no place on the baseball field. In the front office, sure. In the owner’s box, obviously. But not on the field, I tell you!

The most commonly considered fantasy alternative is “Quality Starts” (QS). QS are not perfect, I grant you, and I acknowledged their shortcomings in one of those earlier Master Notes, as well as to a thoroughly bewildered Norwegian gentleman on a long bus ride once. But surely we can all agree that “six-plus innings, three or fewer earned runs” are better criteria for a useful pitcher performance than “happened to be the last pitcher on the mound before his teammates scored a bunch of runs.”

The data in the PQS logs show that 56% of this year’s QS were wins, 30% were no-decisions and just 14% were losses. The combined ERA/WHIP of QS: 1.92/0.91, which compares very well with the decimals in all starter wins: 2.11/0.97. More importantly, the QS are by definition good starts that gave their teams chances to win.

Moving to QS eliminates outright injustices like the April 3rd game where Luis Castillo of the Reds (I know, I know) went seven innings, allowed one run, one hit and four walks, with nine Ks—and lost. Or Jose Berrios, (and the fact that I own Berrios on two fantasy teams doesn’t factor into my chagrin), who went 8 IP, 1 ER, 5 H, 0 bb, 10 K—and lost. Or Mike Leake, by merest coincidence also until recently on two of my fantasy teams: 7 IP, 1 ER, 6 H, 0 bb, 2 K—and lost. Or Joe Musgrove, not on any of my teams—so there!—who in two starts combined, had 13 innings, 1 ER, 10 H, 3 bb, and 11 K—and lost them both, one on an unearned run from a defensive error!

The unfairness—to pitchers and their fantasy owners—was also seen in games like starts from Mike Minor, Kyle Hendricks and Tyler Beede. Each went eight innings, and combined they gave up 15 hits, three walks, 21 Ks and no earned runs. In all three cases, they got no-decisions.

So in these cases, and literally dozens of others, the pitchers were superb, but for reasons entirely out of their control, their fantasy owners got no credit beyond the contribution to ratios and strikeouts.

(Interestingly, Mike Leake, who was on the list of hard-luck guys who deserved wins they didn’t get, is also on the list of guys who got wins they didn’t deserve.)

The best argument against adopting some kind of Quality Start category to replace wins is that it reduces value for non-closer relievers. But why not incorporate some sort of reliever-specific category, like clean relief innings or a Pure Quality Relief kind of thing? (I actually invented a version, but so long ago that I don’t recall how it worked!) That way, maybe we could also get rid of the second-dumbest category: saves.

But don’t get me started. Unless we're on a long bus ride together some time. Seire skal ikke brukes!


Something That Every Baseball Story Seems To Get Obviously Wrong, But Nobody Has Noticed Until Now

The other day I was listening to a CLE game, because I have a couple of CLE hitters and their entire late-inning bullpen except the closer.

The announcer was discussing CLE’s recent hot streak, during which they had climbed from one game under .500 on June 2 to 21 games over .500 on Aug. 4, and he added that they had closed the gap behind division-leading MIN from 11 1/2 games to three.

I thought something didn’t add up. Let me explain:

On June 2, CLE’s record was 29 wins and 30 losses in 59 games. Using the common formula, the announcer said CLE was one game under .500. But a .500 record through 59 games would have been 29 1/2 wins and 29 1/2 losses. So by simple arithmetic, CLE was half a game under .500 (30 losses instead of 29 1/2),  not one game under .500.

Look at the AL_Central standings as of Aug. 4, after 111 games each:

       W    L   Pct.  GB
MIN   69   42  .622   --
CLE   66   45  .595  3.0

By the usual calculation, MIN at 69-42 was “27 games over .500,” while CLE at 66-45 was “21 games over.” Thus since 27-21 equals six, CLE must have been six games behind MIN. But the official MLB standings said CLE was only 3.0 games behind MIN, because the Games-Behind calculation correctly adds one-half game per win and subtracts one-half game per loss. CLE has three fewer in the wins column (1.5 games) and three more in the loss column (1.5 games). And 1.5 x 2 is 3 games total behind.

In terms relative to .500, MIN is 13 1/2 games over .500, CLE is 10 1/2 games over .500, and they are thus 3 games apart.

I rest my case, and I expect the court to side with me. I expect jury nullification, though.


You Thought Jose Ramirez Was Finished? Jose Ramirez Has Other Ideas

Speaking of CLE, it wasn’t that long ago that Jose Ramirez owners were wondering whether to sell low on the season’s biggest offensive disappointments, and be happy to get 20 cents on the auction dollar. Through the first 53 games of the season (June 2), Ramirez was literally one of the worst players among the 138 hitters with 200+ PA in the period, except in SB/650PA, where he's third:

  • HR/650PA: 11, 128th/138
  • RBI/650PA: 45, T133rd/138
  • R/650 PA: 53, 135th/138
  • BA: .206, 136th/138

Ramirez’ skills were actually pretty decent at the time, although his Hard-hit percentage was down from career norms, to 37%. Perhaps as a result, he was being clobbered by a 23% Hit Rate and a 5% HR/FB rate.

Since June 3, however, Ramirez has been reassuring his owners in a huge way. He's still contributed SBs, 11th out of 126 hitters with 200+ PA at 26 SB/650 PA. But he's shot up in the other roto categories:

  • HR/650PA: 35, T37th/126
  • RBI/650PA: 124, T7th/126
  • R/650 PA: 109, T27th/126
  • BA: .286, 47th/126

If we give each of the 126 hitters in the period rotisserie points in the /650PA counting stats and batting average, Ramirez is 11th, behind Fernando Tatis Jr., Ronald Acuna Jr., Christian Yelich, Freddie Freeman, Javier Baez, Yuli Gurriel, Jeff McNeil, Starling Marte, Mike Trout and DJ LeMahieu. Pretty good company.

There are signs of a skills recovery. His Hit Rate for the period was 27%, still below career norm but trending the right way. He pulled the ball in 53% of his batted balls, up nine points from the earlier period, and decreased his IF rate by a third. He exchanged four points’ worth of soft contact for four added points of hard contact, and he made more contact: 11% K rate is down from the earlier 15%.

It’s tempting to build a narrative here to “explain” the turnaround, and if I were challenged, I’d wonder if it had something to do with the whole pregnancy and birth of his child. If all goes perfectly, it can wear on a guy, and any difficulties can be very troubling. Even a joyous outcome means plenty of sleepless nights. Just speculatin'.

It remains to be seen whether Ramirez can maintain this momentum, or if he will fall into a swoon like he did last season. But if you’re a Ramirez owner, and you’re thinking of selling now, ask for more than 20 cents on the dollar. Way more.


More Fun With PQS: Best and Worst Scores, Proving Pretty Definitively That Stephen Strasburg Is A Better Starter Than Clayton Richard

The PQS log lists 281 starting pitchers. The pitcher with the most top scores of PQS-5 is Stephen Strasburg of WAS, with eight. Three more pitchers—Trevor Bauer, Jacob deGrom and Lance Lynn—have seven each. Only four pitchers—Max Scherzer, Charlie Morton, Walker Buehler and Gerrit Cole—have at least five PQS-5s and no PQS-0s.

Speaking of PQS-0s, Antonio Senzetela, mentioned earlier for his win in an eight-run catastrophe, leads MLB eight goose-eggs in his 18 starts. He also has two PQS-5s. Talk about all or (mostly) nothing! Meanwhile, Trent Thornton and Adrian Sampson have seven PQS-0s, and no fives. Having no PQS-5s isn't totally odd—39 pitchers (14%) have none.

The PQS average for all starters with 10+ starts is 2.3. The highest PQS averages are Scherzer at 3.8 and Strasburg and Justin Verlander, at 3.5. The lowest PQS average is Clayton Richard, at 0.7.

Richard has had 10 starts, with none of them scoring higher than PQS-2: two of those, plus three PQS-1s and five PQS-0s. Only one 10+ starter has a higher percentage of PQS-0s than Richard’s 50%: Glenn Sparkman has seven zeroes in his 13 starts, for a 54% rate. Sparkman has a couple of PQS-5s and a PQS-3, however, so his overall average of 1.3 is quite a bit higher than Richard’s.

Click here to subscribe

  For more information about the terms used in this article, see our Glossary Primer.