This is the Quality Start with a sabermetric DIPS twist, and it gets really easy to calculate once you get used to it. I don't think it's the end all or be all, but then nothing really is that. It is, as I like to say, another piece of the puzzle. A dominating start is scored a 4 or 5 and a disaster start is scored a 0 or 1. DOM% is the percentage of starts that are dominating, DIS% is the percentage of starts that are disasters (any start under 5.0 IP is automatically a 0, or disaster).
What's Good and What's Not
From my observations, a DOM at or above the 40% mark is indicative of good pitching; above 50% is great; above 70% is elite. A low DIS is also indicative of good pitching, just look at the table in the link above showing DOM% and DIS% on the axes.
Basically, you want to see a pitcher's DOM% to be over 40% and ideally over 50%, and you want their DIS to be under 20% and ideally under 10%. For example, Johan Santana has a 76% DOM and 3% DIS in 2006 (2.77 ERA), whereas Orlando Hernandez had a 52% DOM and 28% DIS (4.66 ERA), and Adam Eaton had a 31% DOM and 31% DIS (5.12 ERA). Read the link (unfortunately, they removed the article and thus the table is no longer available, sorry), as I noted, there's a nice chart there showing the combination of high DOM% and low DIS%, and there you can see particularly how a low DIS% is so important to a low ERA.
If you had to chose a high DOM% or a low DIS%, pitchers tend to have a lower ERA when you have a low DIS% vs. a high DOM% (obviously if you combine both, you have a much better chance of having an elite pitcher). But I think when the DOM% is high enough, you win more by choosing a high DOM% over a low DIS%, as there are more high quality games pitched overall.
I wholeheartedly recommend buying Baseball Forecaster and learning more about their methods of analyzing baseball. It has been greatly illuminating for me, and if you want to get a taste for it without paying full price, they used to sell their old editions of their annuals on their website for half price or less (plus shipping); but that was before he sold the company off, and I haven't checked recently.
Giants Starters' PQS for 2016 SeasonMadison Bumgarner- (82% DOM, 0% DIS; 9:0/11): 2, 5, 2, 5, 5/5, 5, 4, 5, 5, 5/
Matt Cain- (44% DOM, 33% DIS; 4:3/9): 4, 0, 0, 3, 3/0, 5, 5, 5, X/
Johnny Cueto - (82% DOM, 0% DIS; 9:0/11): 4, 4, 5, 3, 5/3, 4, 4, 5, 4, 5/
Chris Heston - (0% DOM, 0% DIS; 0:0/0): //
Jake Peavy - (36% DOM, 27% DIS; 4:3/11): 3, 0, 2, 4, 0/3, 3, 5, 0, 4, 4/
Jeff Samardzija - (73% DOM, 0% DIS; 8:0/11): 3, 4, 3, 4, 4/5, 4, 4, 5, 5, 3/
Albert Suarez - (0% DOM, 0% DIS; 1:0/2): //
X - Cain had a disaster start, but was dealing when hamstring went out, not counting in PQS stats
Giants Season overall - 64% DOM, 11% DIS out of 53 games counted (34:6/53)
Giants Month of April - 48% DOM, 16% DIS out of 25 games counted (12:4/25)
Giants Month of May - 79% DOM, 7% DIS out of 28 games counted (22:2/28)
The month of May for PQS was stupendously brilliant! I don't have time to go through past stats, but I'm 99.9% sure that 79% DOM is the best month ever! Mainly because Zito was a big part of the past rotations and he was rarely this good at delivering the DOM starts, then Lincecum in recent years was up and down.
Bumgarner, as he has in recent years, get into high gear after early season issues (seems to happen most years than not; he actually said that he was still not feeling it into late May) and led the staff with 6 DOM starts out of 6. Both Cueto and Samardzija followed closely with 5 DOM starts. And both Peavy and Cain delivered as well, with 3 DOM starts.
And the great thing about throwing so many DOM starts is that there is fewer starts for the DIS starts that are basically guaranteed losses (the Giants were 0-6 in DIS starts to EOMay, 0-2 in May). There were only two in May, leading to a stellar 79% DOM/7% DIS, which led to a stellar 21-8 record in May, in spite of the offense only averaging 4.07 runs per game. This zoomed the Giants into first place, until they took over on May 15th and kept on building since. In 8 days, they pushed it to 4.5 games ahead, then was amazingly matching the Dodgers to the end of the month, staying exactly 4.5 games ahead, win or lose (it helped that they went 5-2, as well).
With the great DOM came great stats. Bumgarner had a 1.05 ERA for May, with a 3.75 K/BB, 9.5 K/9 and 0.961 WHIP. Cueto was next with 2.03 ERA, 4.33 K/BB, 7.9 K/9, 0.992 WHIP. Samardzija started having results to go with his good PQS stats, 2.08 ERA, 5.00 K/BB, 8.3 K/9, 0.969 WHIP.
Cainer was finally starting to feel back to pre-Perfecto form, with 3 DOM in 4 starts, compiling a 3.38 ERA, 4.25 K/BB, 5.7 K/9, and 1.275 WHIP. Still giving up more hits and homers than usual, but keeping the walks down helped, clearly. Too bad he had the injury hiccup with the strained hamstring, he was starting to feel it good.
Peavy was still getting crushed when he was on, in spite of his good PQS for the month. He had a 4.73 ERA, 2.36 K/BB, 7.2 K/9, and 1.268 WHIP.
Still, great month for the starters overall, stupendous, really. They collectively pitched as if they were the best pitcher in baseball, pitching in every start, with that 79% DOM and 7% DIS. And we should have been even better in the win column. Had we gotten some better offense, that would have really put some distance between us and the NL West. Though, really, had we gotten better bullpen pitching (3 blown saves), specifically Casilla, we could have been 24-4 instead. But this is getting picky, all in all, the Giants starting pitching was great in May.
May 2016 Comments
I may keep it shorter, since June is basically half over as it is (but I yam whut I yam...). As bad as April was (relatively, by Giants past standards), May was that good. The offense that saved them in April with 4.96 runs scored per game, helping them get to 12-13 in spite of all the bad pitching in the back of the rotation, was stagnant, averaging only 4.07, and it won't get much better with Hunter Pence missing big parts of the coming months (hopefully to return by start of August).
The biggest difference between the months is that while only Span and Duffy did not hit well in April, more did poorly in May, with a bifurcation in the offense in May. While Tomlinson, Pence, Belt, Posey, Crawford, and Span were hitting well, Parker, Blanco, Duffy (still...), Gillaspie, Panik, Williamson, Brown, and Pagan were not. Parker at last was hitting for power, so his OPS was .686, which is where Span and Duffy roughly was in April. All the rest from Blanco on down, ranged from Blanco's .628 OPS to Pagan's .439 OPS (probably already showing the effects of his hamstring issues, half of what he did in April). And none of the OF options has done well enough to justify given them more AB's.
But when a defense (pitching and fielding) can keep runs allowed to the 3.24 average in May, it covers up a lot of offensive ills. It allowed the team to go 7-0 in one-run games, pushing the team up to 15-7 in one-run games.
And the good news is that May is greatly improved over April and look, at the moment, closer to the true talent level of the rotation. Peavy is reaching his regular second half peak that he's been doing for us the past two seasons. Cain was dealing until his hamstring, but Suarez did well in taking two starts for Cainer, keeping the line moving in the rotation. With Bumgarner, Cueto, and Samardzija looking relatively dominant, all we need is one of Peavy and Cain to do well in order for the Giants to continue winning more than they lose. In addition, nobody on the team is striking out a ton, so the poor OPS for many of these players could just be a case of BABIP hurting them. In particular, Duffy, Panik, and Pagan all have BB/K ratios above or very near to 1.0, and generally those hitters hit for a much higher BABIP.
As noted last month, the Giants were below Pythagorean last month, and they regressed to the mean by playing above Pythagorean in May. They should have been 17-12, but ended up 21-8, 4 games above. And in April, they should have been 13-12, but ended up 12-13.
Most sabers would then say that the Giants should regress back to .500, because of May, but as my research showed, Bochy has been able to maintain a +4 games above .500 record over his managerial career, until the past two seasons. So 2016 could be his career-wide regression to his talent mean, meaning 15-7 (+8 and higher has been attained by Bochy in nearly 40% of his career managerial seasons) could just be Bochy managing to his talent level.
Lincecum Era Over (For Now)
Unfortunately, the Angels had a rash of injuries in their starting pitcher, so they offered Lincecum a starting job, roughly in the $4M range (but pro-rated), so he didn't get a lot of money, but that wasn't his goal, he just wanted the opportunity to start.
I still think that Lincecum will be very good when he come back and have a very good season, setting him up nicely for a free agent contract after the season. Will he stick by the Angels because they gave him the opportunity?
Perhaps, but, at least in the past, when he was an amateur and in his early years in arbitration, his agent would try to get what they thought he was worth. He could have been an Indian all these years had they given him the $1M he wanted when they drafted him the year before (only offered $400K). And he held out for $200K more from us when we drafted him. He and the Giants had the widely different arbitration numbers during those years.
The Giants will see Peavy's contract ending this season, opening up a spot in the rotation. Nobody so far is making the Giants management say that they should be starting in the majors instead. Blackburn had his early struggles, though Giants Prospect Talk noted that high altitude was a factor in that. And he has an 0.66 ERA since returning from the majors. And Cain's contract ends after 2017 (plus Cueto most probably opts out), so openings in the rotation are available, should the Giants decide to pursue Lincecum in the off-season.
Will they? First, we have to see how Lincecum does, but if he does as well as I think he can, he can probably get a pretty big one-year contract (I just don't see anyone giving him a long-term big money contract based on one year, plus he seems to favor shorter terms than having the weight of a huge contract on his conscience). And the Giants probably would be very fine with such an arrangement, going year to year. But so would other teams, if he's as good as I think he is.
I would hope the Giants open up contract discussions with Bumgarner and Cueto before it gets critical. I was very big on pursuing Cueto whom I think is one of the best pitchers in MLB. And, of course, Bumgarner is just amazing.
ReplyDeleteAlso if Lincecum can show he can pitch in the majors, I'd be happy if they bring him back.
I agree about Bumgarner, but I don't want two humongous contracts in the rotation like that, he'll be in Greinke territory, like Madison. Get two years of great pitching and move on. Shark hopefully continues to do well for another 3years, plus hope that one of Beede, Bickford, Blackburn, other can take over as third ace of staff. And maybe Cain returns to that status. And I think Timmy is capable, I'm just afraid of how short that could be.
DeleteI have a couple of comments, the first one brief: can you add a sentence or two to the intro above, defining what counts as a DOM and what as a DIS? Can these be given equivalents in Game Score?
ReplyDeleteThe other matter is more complicated, and is a complement to your way of evaluating starters by DOM/DIS. How does one evaluate bullpens? (I tried to raise this issue on MCC, objecting to some glib comments from Grant, but his followers were more interested in discussing, among other things, how often they defecated. I'm not kidding.) The usual measures we use for starters, such as ERA, FIP, and WAR, don't work well; and different relievers and different 'pens have markedly different leverage situations, which makes comparisons of effectiveness difficult.
Do you, as a skilled and creative analyst, have measures you use to weigh the effectiveness of bullpens? Strand rate would need to be included, WHIP with an allowance for intentional walks, perhaps; also Runs Allowed. I'm sure that if one wants to consider how badly the Giants need to beef up their bullpen--top prospects for 31-year-old Andrew Miller?--one needs to have a clear sense of how the 'pen is doing; but are there clear, statistical means for doing that?
I guess you forgot, since I keep the text on top the same, but the link to Baseball HQ is suppose to have the definition of DOM and DIS there.
DeleteGame Score does not technically have an equivalent, per Bill James creation, but a saber compiled stats empirically and labeled GS above a # to be a Win and below, a loss. Basic same structure as PQS, win=DOM, loss=DIS, tie=what I call MID, because they never defined one.
I don't like putting the definition in my post, since it is their creation, but I have put it in comments before, so will do it again.
DOM= when PQS is 4 or 5
DIS = when PQS is 0 or 1
PQS is a simple rating system that is saber metrically aligned. Get automatic 0 if you don't reach at least 5 IP. I leave it out if the pitcher left due to injury and would have otherwise not gotten a 0. Add 1 if 6 IP. Add 1 if hits are less than or equal to IP. Add 1if K>=2 x BB. Add 1 if K>=IP-2. Add 1if less than 2 HR given up.
So it rewards IP min of 6. If the pitcher does not give up too many hits, basically 1.000 HIP. If the pitcher does not give up too many walks, i.e. want SP to have K/BB ratio >= 2.0, which was above average during the 2000's. If the pitcher strikes out a good number, their study showed that you want above 6.0 K/9, so this is approximation. Lastly if you don't give up too many HR.
As I noted in my last post on PQS, Game Score is not saber designed based on recent rules of thumb for good pitching. James seems to have a brilliant eye for getting at the gist of things, so I would say that GameScore is good at rating how well the pitcher pitched in terms of results, but PQS is better at ignoring the runs results, which are skewed by varying degrees of luck in scoring, and looking strictly at the pitching mechanics results. To me, and I've never saw anyone try to describe them in any way, GameScore is like ERA and PQS is like FIP. So each has value, as long as you know what you are dealing with, as with any metric used as a tool.
DeleteI like PQS because it can help me see when a pitcher is doing well, but just getting unlucky giving up a lot of runs, which hurts Game Score. That I don't always see from a pitchers line, it gives a nuance. Whereas I can see from a pitchers line that he had a good or bad Game Score, the only question is the degree. That's interesting but I feel it doesn't add much above what I get from looking at his line.
Also, a bad game could really screw up your seasonal total in GameScore, like with ERA, but a DIS only affects that one game, you have to screw up multiple starts to really affect your DOM% and DIS%.
So I prefer PQS.
There are no accepted rules on rating a reliever, or providing a way to compare two different ones, that I have seen. I think ERA or FIP is acceptable when looking at a group or over a career.n WPA looks good too, but as you note, there are issues. I like seeing what their inherited runners scoring rate is, obviously the lower the better. I think having high K relievers is a great way to go, less chance for BABIP God to smite thee at the wrong time. And one with very high K/BB ratio, love guys coming up with 4+ K/BB ratio, like Strickland, Law, Okert, Osich. This allows me to be OK with relievers with high walk rates, as the K's make up for the walks by limiting BIP and thus total hits.
DeleteAs far as trades go, you basically never want to trade any prospects who looks to be a good starter just to get a reliever. Thus you need to rely on your scouts to make the right judgement on who to trade. So far, Sabean et al has been very good at limiting losses of prospect talent in trades.
DeleteIt would make me very nervous to see the Giants trade for a top reliever by giving someone who looks to be good. Heck, nervous for any starter too. At least until the prospect don't look good any more.
I remember regretting Caruso, but he just fizzled. So started my lessons that prospects aren't really that good at MLB level unless they exceptional in the minors. Then Crawford made things tough again.
It is very much an art, I'll have opinions but if the Giants let go of someone, I assume I'm just missing something, as I don't have any scouting sources. So far, for me, so good, I have not disagreed with the trades the Giants have made, for the most part.
And Miller I would have issues getting, not for good reasons, but because we could have had him, I remember the Giants signing him, then we lost him not that nuch later, for some reason. :)
Personally I think our bullpen is fine for now, any big trade you will have to pay thru your nose. I would rather supplement later, when teams are more willing to give up, that's how we got Lopez and Ramirez. I also prefer giving opps to our prospects, we have so many in the upper minors, just bring them up, see what we got, like they did with Blackburn, starters can get feet wet by pitching in relief first. Beede could be ready by Sept.
Thanks for the lucid explanation as to PQS vs Game Score, and for the further definition of PQS, for which I needed a convenient reference point, as might other continuing readers of your blog, and certainly any new readers.
DeleteI too trust the Giants FO. But I am interested in evaluating the call for upgrading the bullpen this season, of which I doubt the pressing need, and I dislike not having a reasonable statistical method for evaluating any potentially evaluable and significant aspect of baseball.
Sorry, I know you trust the Giants FO, was just talking from my viewpoint.
DeleteI would love to be able to evaluate relievers better. Until someone solves the issue with leverage, it's very much an art, not that evaluating other players are an exact science yet, but closer than relief pitchers.
Clearly, though, a playoff team needs to have a good bullpen overall. Tom Tippett researched this at his old site before joining the Red Sox, and found that teams needed to not only be good offensively and in starting pitching, but also high in the ranks of relief, as well.
I like wpa for evaluating relievers, it weights situations differently according to how important they are in a game and it also assigns credit or blame realistically when one pitcher replaces another in the middle of an inning.
ReplyDeleteThanks. My reservation about WPA is that it depends heavily on leverage, and that in turn depends on managerial decisions about whom to use and how, and on the performance of whom one follows on the hill. Comparing the 'pen of Team A and that of Team B becomes difficult. That all doesn't mean that WPA may not be the best way to measure bullpen effectiveness, just that it doesn't seem ideal, particularly in isolation.
Delete