Wednesday, July 11, 2012

It's all about the pitching

Baseball-Reference measures pitching using ERA+, which compares the league average ERA to a team or individual ERA. Thus, an ERA+ of 120 roughly means that the league is 20% worse (gives up 20% more earned runs per nine innings) than the pitcher or staff. FanGraphs uses ERA-, which flips the fraction over and compares an individual or team ERA to the league average. Thus, an ERA- of 80 would indicate a pitcher or staff that is 20% better (allows 20% fewer earned runs per nine) than the league. In both cases, a score of 100 would indicate unity, or league average. Both stats claim to have a park factor in the final result that adjusts for, say, playing in Fenway vs. Petco. Both are handy thumbnails for evaluating pitchers and pitching staffs.

I made a graph of the Giants team ERA+ and ERA- for the seasons 2009-2012. I treated the interim numbers for 2012 the same as the full-season numbers, just to get a quick snapshot of how the current guys compare. Madison Bumgarner actually made his debut in 2009, throwing ten innings over four games (one start). Tim Lincecum was the star of that squad, and both Matt Cain and Barry Zito had 33 starts apiece. Brian Wilson, Sergio Romo, Jeremy Affeldt, and Brad Penny were also on that staff. In 2010, The Bumbino had 18 starts and supplanted Barry-Z on the post-season roster. Santiago Casilla, Guillermo Mota, and Javier Lopez joined the fun. Last year Ryan Vogelsong emerged and Jonathan Sanchez pitched his last game as a Giant. Take a look:

As a team, the Giants ERA+ for 2009-20012 goes 120, 117, 110, and 97. That's 1st, 1st, 3rd, and 10th in the NL. Their ERA- goes 86, 85, 86, and 93, good for 2nd, 1st, 2nd, and 5th. I don't know why B-R rates the Giants so poorly compared to FanGraphs, but I suspect it has to do with the mysterious "park factors." Regardless, it is clear the staff is not as elite as it once was. We all know the culprit, of course, but Tim Lincecum gets some help from his bullpen, too. The loss of both Willie and WillyMo certainly hurt. The Giants are blessed with excellent starters and a talented relief corps. But team-wide they are much more prone to giving up runs than in the previous three seasons. Another way to look at this is to just take raw "runs allowed" numbers and compare them to the league average "runs allowed." The RA/lgRA ratio should be 1.00 for a "league average" team, eh?

2009: 611 RA, NL average 727; RA/lgRA 0.84
2010: 583 RA, NL average 705; RA/lgRA 0.83
2011: 578 RA, NL average 673; RA/lgRA 0.86
2012: 346 RA, NL average 368; RA/lgRA 0.94

No surprises, really. We know the staff is giving up more runs this year. The question is, will this trend toward average-ness continue? And will that hurt the team's chances? If the answer to the second question is "yes" (and I think it is) then we have to hope the answer to the first question is "no" (and I think it is). The Giants will find a fresh arm or two for the 'pen, and the laggards and slackers will pick it up. This I believe. Random events (injuries, for example) can derail the best-laid plans, so we'll just ignore them, OK?

The Second Half is upon us! I feared the Baseball Gods would punish Giants fans for stuffing the ASG ballot box, instead they blessed us for it. Let us hope they continue to bless the rest of the season. Perform your ablutions and make your sacrifices, my friends, the dark forces must be propitiated!




nomisnala said...

The 2012 giants sure can give up the unearned runs this year.

M.C. O'Connor said...

Yeah, the shockingly porous defense has not helped the pitchers. I hope that improves in the second half.

Zo said...

My question is whether this same sort of analysis can be applied to starting vs. relief pitching, or further broken down by mid-relief vs. closers? Relievers can be easily tinkered with, but starters, not so much. My gut says it is relief, maybe skewed by Castilla's two blown saves just recently. He has 21 saves, half of a good closer's year. We should have three starting pitchers with 10 wins by now. A win doesn't always indicate a pitcher's strength, but then, era (+ or -) doesn't always indicate a team's success. But three 10 w guys by the AS break does indicate that we have had a lead while our starters were in there a significant part of the time (and, I'm betting, a lot more than the last couple years). This just reinforces my gut feeling that our trend away from dominance this year is relief, not starting. Of course, I'm too lazy to pull out the stats, to determine this, but maybe someone else isn't.

M.C. O'Connor said...

ERA and ERA+/- are not the best stats for relievers. They can usually tell you who is doing well, but one or two bad performances can really skew the numbers and make a good pitcher look worse. Relievers just don't rack up enough innings. It's like evaluating a hitter's season after 100 PA. That being said, I've no doubt, other than Tim, the relief pitching is the main culprit. Besides, we all know that ERA is one of those stats that's really good at finding the outliers (who is really good and who is really bad), and not so good for evaluating guys (starters) in the middle of the pack. It is too dependent on environment, relief, random variation, etc. It's like batting average in that regard. I mean is a 3.33 pitcher actually better than a 3.74 pitcher? Not necessarily, you'd have to look at context. But a 2.33 pitcher is clearly doing better than a 4.74 pitcher.

I think the way to evaluate the 'pen would be to look at things like BB/9 and K/9. And things like are ground ball specialists like Affeldt getting ground balls? Are guys like Lopez getting lefties out? Intuitively, I'd say "no" but I have not done the looking. And there's no doubt the Giants "garbage time" guys have been terrible. A game they would have lost 6-1 winds up 10-1 because it is Loux instead of Mota, for example. (I've been hard on Kontos, but I think they may have a good arm there and he may become a key guy down the stretch.)

I'm not sure Wilson would have been better than Casilla so far, but having Wilson would mean that Casilla could get some big outs in setup and pre-setup situations, giving the team, in effect, two Romos. Kind of like when they had that great Ramirez/Casilla combo in 2010. So losing the closer has indeed hurt the club.

I expect a fresh bullpen arm will be acquired before the deadline. Huston Street is available, but the Padres will want a lot, and the Giants would probably regret giving up a key prospect, so I doubt they'll even consider it. But boy he'd really make the 'pen dynamite and take a lot of pressure off the other guys.

Zo said...

The fresh bullpen arm is Penny. Doesn't mean that they should stop there.