Swapping Brandons

On Monday night, the Blue Jays and Mariners hooked up on a trade for the second time this offseason --- the previous week, they had been involved (along with the Blue Jays and Athletics) in the Roy Halladay-Cliff Lee blockbuster. This trade featured only three players and obviously didn’t get as much publicity. The Mariners sent 2006 first-round pick Brandon Morrow to the Jays for reliever Brandon League and 20-year old minor league outfielder Johermyn Chavez. The big-leaguers involved have quite a few similarities --- both Morrow and League are hard-throwing, right-handed pitchers with great stuff who have not yet been able to have sustained success. Also, they’re both named Brandon. Just for fun, I would like to look at both of these pitchers with PITCHf/x. Let’s start with Morrow.

Brandon Morrow

Morrow was the 5^th overall selection in the 2006 amateur draft. In his three major-league seasons, Morrow has both started and relieved, compiling a 3.97 ERA, 9.29 K/9, and a 5.83 BB/9 over 197 2/3 innings. He’s a four-pitch pitcher, and there’s nothing “soft” in there. Here’s his stuff.

Brandon Morrow Pitch Data

Pitch	Average Speed (mph)	Max Speed	pfx_x	pfx_z	spin_angle
FF	94.91	99.7	-3.31	11.35	196.07
SL	86.64	90.5	2.65	0.79	109.50
CH	88.42	91.3	-7.05	6.87	226.35
CU	81.40	83.6	2.54	-5.21	69.07

Hard fastball, hard slider, hard changeup (it’s more or less a splitter), hard curveball. The changeup (or splitter)/fastball and slider/curveball combinations seem to run together sometimes, so there may be a few changeups classified as fastballs (or vice versa) or curveballs classified as sliders (or vice versa). The stuff itself is impressive, as are the whiff rates (especially for the fastball and slider):

Pitch	Pitch#	Wide Zone%	Swing%	Whiff%
FF	874	52.6%	44.6%	17.9%
SL	204	46.6%	54.4%	47.7%
CH	125	52.8%	48.8%	11.5%
CU	42	21.4%	31.0%	23.1%
	1245	50.6%	46.2%	23.1%

Morrow has expressed the desire to start in Toronto, and the Blue Jays seem willing to honor that request. His average fastball velocity dropped from 95.79 in the bullpen to 94.36 in the rotation. Although that’s a significant drop, 94.4 mph isn’t too shabby ---- only four qualified starters in 2009 had faster fastball velocities.

Brandon League

Like Morrow, League has the ability to miss bats. Before I look at that more closely, here are the graphs and charts for League’s repertoire.

Brandon League Pitch Data

Pitch	Average Speed (mph)	Max Speed	pfx_x	pfx_z	spin_angle
SI	95.37	98.8	-11.41	2.98	255.51
FS	85.47	88.3	-5.89	-2.75	292.90
SL	89.55	91.4	-2.47	2.39	232.36

There were a few stray pitches that looked like kind of like sliders, so I labeled them as such. But for the vast majority, League works with sinkers and splitters, both of which have some serious vertical “drop.” The movement and velocity combination is enticing enough, and, as I mentioned earlier, it led to some serious whiffing:

Pitch	Pitch#	Wide Zone%	Swing%	Whiff%
SI	720	64.0%	51.4%	10.3%
FS	370	30.8%	49.5%	68.3%
SL	11	27.3%	36.4%	75.0%
	1101	52.5%	50.6%	29.8%

The 29.8% total whiff rate is spectacular, and the 68.3% mark on the splitter is beyond ridiculous. in fact, his offspeed offering was the most unhittable pitch in the majors last year. When he wasn’t avoiding contact, League was getting groundballs at a 55.7% rate (actually down from 66.7% in 2008, but still excellent) --- but he didn’t get great results with his 4.58 ERA. However, he did have a 3.80 tERA and 3.58 FIP last year, and if he can retain his peripherals, he could improve his numbers vastly due to the ability of the Mariners’ defense and the .... pitchers'-ballpark-y-ness (?) of Safeco Field.

To some, this deal may seem like a steal for the Blue Jays --- a 2006 first rounder with ace-type stuff for a reliever and a 20 year old outfielder? And who knows; if Morrow turns into the Jays’ ace for years to come and League and Chavez fizzle out, it will be a steal. But we must remember that Morrow has been an injury risk in his time in the majors, and that on paper League is a great fit for the Mariners' back-end bullpen due to his pitching style. And if Chavez becomes a useful player or trade chip, the deal could become that much better for the Mariners.

Gameday PITCHf/x data is from MLB Advanced Media; it can be easily accessed via this tool. Other statistics are from Fangraphs and Baseball-Reference.

Hughes' postseason struggles

A few weeks ago, I used PITCHf/x to look at Phil Hughes’ interesting 2009 season. I was asked to do a follow-up on Hughes’ less-than-stellar postseason. I’d like to share a few of the things I found.

Obviously, Phil didn’t get the kind of results in the postseason as he did in the regular season:

Regular season: 1.40 ERA, 1.83 FIP, 11.4 K/9

Postseason: 8.53 ERA, 4.83 FIP, 9.95 K/9

What’s strange is that while he got absolutely tattooed, he still had a high strikeout rate. Here is his “stuff” (on the left is from his regular season bullpen stint, on the right is from the postseason):

Pitch	Average Speed (mph)	Max Speed	pfx_x	pfx_z	spin_dir
FF	94.64 94.32	97.3 96.3	-4.98 -4.54	10.33 10.40	205.80 203.47
FC	89.26 88.92	92.6 91.9	1.10 1.17	5.27 5.68	166.79 167.33
CU	77.49 77.53	80.4 79.1	7.27 6.25	-7.86 -9.44	42.78 33.73

Hmm. That’s not much different. It would make sense that if a pitcher all of a sudden starts getting hit hard, it would correlate with some sort of change in raw stuff, but that doesn’t seem to be the case here. There’s a slight loss in velocity, but I don’t see how that could have much of an effect overall. Below are the pitch results. Swing% is swings / total pitches; Whiff% is swinging strikes / swings; Wide Zone% is the vertical strike zone set by the PITCHf/x operators and a two-foot horizontal zone (one foot off of the center of the plate); Chase% is swings out of the wide zone; Watch% is takes inside the wide zone.

Pitch	Pitch#	Pitch%	Swing%	Whiff%	Wide Zone%	Chase%	Watch%
FF	554 87	66.3% 62.1%	52.2% 42.5%	27.0% 32.4%	56.7% 48.3%	26.0% 35.1%	31.8% 42.9%
FC	107 23	12.8% 16.4%	56.1% 70.0%	33.3% 12.5%	65.4% 65.2%	21.7% 6.3%	32.9% 20.0%
CU	166 30	19.9% 21.4%	34.9% 63.3%	19.0% 42.1%	44.0% 56.7%	25.9% 42.1%	41.1% 35.3%
	830 140	99.4% 99.3%	49.0% 51.4%	26.8% 30.6%	55.1% 52.9%	25.3% 30.6%	33.5% 36.5%

I’m still not seeing what could cause that much of a decrease in effectiveness. The cutter is the only pitch that didn’t see an increase in whiff rate from the regular season. There is a decrease in Wide Zone% --- I’ll get to that in a minute. Before I get to some thoughts on what might’ve been wrong, I’d like to show a final set of numbers comparing his regular season and his postseason. SLGCON is slugging on contact, or total bases / balls in play. Runs Saved is linear weights based on pitch outcomes (ball, strike, single, double, strikeout, etc); RSv/100 is Runs Saved per 100 pitches. Expected Runs Saved is calculated the same way as Runs Saved, except that it is based on batted ball types (groundball, flyball, line drive, pop up) instead of outcomes; xRSv/100 is the rate stat.

Pitch	Pitch#	SLGCON	RSv/100	xRSv/100
FF	554 88	.406 1.111	2.57 -3.88	1.82 -0.23
FC	107 22	.263 1.000	4.56 -1.61	2.83 -0.76
CU	166 30	.296 .429	0.61 2.50	-0.99 1.69
	830 140	.357 .842	2.40 -2.14	1.37 0.13

So it’s clear that something happened. Hughes turned from exceptional to horrible, although you wouldn’t be able to tell from his peripherals. I had a few thoughts as to what might have been Hughes’ problem in the postseason, so let’s get to those.

The first thing I’d like to look at in terms of graphs is his pitch location on hits in the postseason.

Nothing earth shattering here; it looks like he hung a few curveballs and left a few fastballs too low.

Moving on! One thing noticeable in the pitch results is that Hughes had a lower zone percentage, especially with his fastball, in the postseason. So, is it possible that he was falling behind in the count and then getting hit hard after falling into “hitters’ counts”? Let’s take a look:

Stint	FPS%	Ahead%	Behind%
Regular Season	62.4%	46.9%	14.4%
Postseason	60.6%	45.5%	12.1%

That doesn't really explain a whole lot. In addition to falling behind hitters at a similar rate as he did in the regular season, he maintained an excellent first-pitch strike percentage. So that theory is essentially debunked.

In his postseason games, I remembered seeing him “nibble” at the strike zone more than he did in the regular season, and I wondered if what I remembered seeing had any bearing on a) whether this was actually happening and b) if this had any effect on his performance (for example, his “nibbling” on the corners might’ve cost him some close calls from the umpire). I figured I could test this out by taking the average distance of each pitch from the center of the plate, and the mean that I got would be the “Nibble Score.” In order to filter out balls way out of the zone, I used the wide zone plus three extra inches on each side of the plate. Also, for here, I only used fastballs. This next graph shows the Nibble Score in a variety of different count situations, dictated by run expectancy. Negative numbers indicate a count favorable to the pitcher; positive numbers indicate a count favorable to the hitter. The count-based run expectancy is 0 on the first pitch of an at-bat.

The sample size is small for the postseason stint, so I’ll let you draw your own conclusions from the graph. There doesn’t seem to be that much of a difference between the overall nibble-ness (0.55 in the regular season, 0.58 in the postseason), in which I’m working with a larger sample size, so I’ll put that theory to rest now.

The last thing I wanted to look at (for now) was Hughes’ release point by pitch. It’s pretty clear that the curveball is coming from a higher slot, and it makes me wonder if hitters were able to sit on his fastball after seeing the lower release point. But still, the curveball is coming from a bit higher in the regular season plot, and he obviously had great success. For what it's worth, here are the charts.

So, I couldn’t find any statistic that gave any clear answer as to why Hughes struggled so mightily during the postseason. He was getting ahead on hitters, he had good velocity, and he had the ability to miss bats. Because of this, I’m not overly concerned about Hughes. However, I looked at some video of the postseason earlier today, and in the (small) sample that I looked at, it seemed as if Hughes was missing location within the zone --- Jorge Posada had to move his glove a lot. This isn’t quantifiable by any metrics we have now, but that could be the most important thing --- pitching to the situation. Although we can use PITCHf/x to give us a good indication of what happened, it can’t tell us what the pitcher intended to have happen, and I think that in this case, this was what burned Hughes the most.

Gameday PITCHf/x data is from MLB Advanced Media; it can be easily accessed here and here. Other statistics are from Fangraphs.

My personal thoughts on the Curtis Granderson trade

Yesterday, the Yankees, Tigers, and Diamondbacks pulled off a blockbuster trade that sent Curtis Granderson to New York and an assortment of young players to Detroit and Arizona. As a Yankee fan, I was at first skeptical of the trade, because I was wary of Granderson’s production last year and I wasn’t that thrilled with having to give up four young players (of varying talent level). But the more I’ve thought about it, the more pleased I’ve become with this trade. Even though Granderson’s average was down last year, he did hit 30 home runs. And Yankee Stadium probably won’t hurt a lefty who pulled all but three of his home runs last year. I also don’t think the Yankees necessarily need to hit him at the top of the order, where he did the bulk of his work last year. Especially with the power he’s shown, he could benefit from hitting 6th or 7th in the offense-rich Yankee lineup. Granderson can also run, which is a plus for a Yankee team that is (once again) getting up there in age. He was a 20/20 guy last year and in 2007, and there’s no reason to believe that he can’t duplicate that. The last thing I’ll say about Granderson doesn’t have to do with his game on the field. He’s viewed as a great guy and a great teammate, and he has served as an ambassador to Major League Baseball. I’m not saying that it’s a good idea to acquire players for their personality traits, but especially in a big city like New York, being a good guy can never hurt.
How about the rest of this monster deal? The Yankees shipped Phil Coke and prospect Austin Jackson to the Tigers and Ian Kennedy to the Diamondbacks. Originally, the deal included a fourth prospect, lefty reliever Mike Dunn, who was supposed to go to the Tigers. Talks seemed to pick up when Dunn, who is 24, was removed from the package. Dunn’s got good stuff and a lot of velocity, but has a hard time finding the strikezone (as we saw in his brief stint with the Yankees last September); still, it seems like the Yankees don’t want to give up on him. Coke, Jackson, and Kennedy is certainly a fair lot for Granderson. Although I was a Phil Coke fan last year, he is certainly replaceable member of the bullpen (maybe even with Dunn next year, if he can keep his control problems in check). Jackson was highly touted and was hyped-up by the Yankee brass, but his stock seemed to have fallen as his lack of power became more pronounced. I’m certainly fine with turning him into Granderson. As for Kennedy, he had pretty much had nowhere else to go with the Yankees, as they have plenty of pitching depth.
The other part of this deal was a swap between the Tigers and the Diamondbacks: Edwin Jackson for Max Scherzer and Daniel Schlereth. The Diamondbacks' part in this deal is debatable, but it’s certainly good for the Tigers. In Scherzer, they get a young rotation arm with good stuff, and in Schlereth, they get a young bullpen arm with good stuff. Although Jackson had a fine season last year, his low strikeout rate and high walk rate point to a likely regression in terms of ERA next year. Then again, he’ll be returning to the National League, which, along with the double play, is a pitcher’s best friend. Also, Jackson is still relatively young, he throws hard, and based on stuff alone it’s reasonable to believe that he can be an above-average major league pitcher. Still, I think that the price of Scherzer and Schlereth was a bit much for Jackson. In Kennedy, the D-Backs get a potential back-end guy behind Brandon Webb (if healthy), Dan Haren, and Jackson.
Overall, I think that the Yankees are the clear winners in this trade. Acquiring Granderson was important in making the team better, and the price they had to pay was not that steep (they didn’t have to part with Phil Hughes or Joba Chamberlain). The Tigers are clearly in salary-shedding mode, and getting Coke, Jackson, Scherzer, and Schlereth is a good start for a team that’s trying to save money and get younger. The Diamondbacks are the losers of this deal in my mind, but if Kennedy and Jackson step up, they could have a dynamic rotation in 2010.

Yankees-Tigers-Diamondbacks blockbuster in place

The Yankees are about to acquire Curtis Granderson in a massive three-team, seven-player deal. The Yankees are sending lefty reliever Phil Coke and top outfield prospect Austin Jackson to the Tigers; the Yanks are also sending Ian Kennedy to the Diamondbacks. The Diamondbacks will send pitchers Daniel Schlereth and Max Scherzer to the Tigers, and the Tigers will send starter Edwin Jackson to the Diamondbacks. Jon Heyman says the deal is pending medicals.

You have reason to be excited

Although the offseason is long and dark for baseball fans, the most exciting part of it is now upon us --- the Winter Meetings. The braintrust of each Major League team will congregate in Indianapolis this week, and we’re sure to see some action. At last year’s winter meetings, CC Sabathia, A.J. Burnett, and Francisco Rodriguez all signed contracts, and Mariners’ closer J.J. Putz was shipped off to the Mets in a massive three-team deal. Although there may not be that much action in this year’s Meetings, which run until Thursday, there are certainly a few players to keep an eye on. Here are four that I think will draw a lot of attention.

4. John Lackey

Lackey has spent his whole career with the Angels, compiling a 102-71 record since his rookie season in 2002. He’s not the most dominant guy on the market (more on that later), but he has a good track record and has been fairly durable, posting 200+ inning seasons from 2003-2007. However, Lackey was hit with injuries in both 2008 and 2009 --- including an elbow injury last year, which is a red flag for general managers. In October, Sports Illustrated reported that Lackey would demand at least the $82.5 million that A.J. Burnett received from the Yankees, but I think he may have to settle for less due to his recent injury history.

3. Jason Bay

Bay made a name for himself in Pittsburgh, winning Rookie of the Year in 2004. After being traded over to Boston in the 2008 season, he came into the national spotlight much more so than he did while playing for the Pirates, and he will certainly be a prized commodity this offseason. His 36 home runs in 2009 were third in the AL, and he finished seventh in the AL MVP voting. Bay is a great player, as he can hit for power (.519 career SLG) and earn his walks (.376 career OBP). The strikeout totals (162 in 2009, fourth in the AL) are high, but I doubt that teams will be scared away by that very much, especially considering Bay’s overall hitting prowess. Bay’s agent expects his client to remain in contact with the Red Sox, and given Bay’s performance (he’s gotta love that Green Monster) and good relationship with the fans, I would say that there's a good chance that he stays in New England.

2. Matt Holliday

I think Holliday will be the most sought-after position player this offseason. Although his .938 OPS over his five-year Colorado tenure has been met with skepticism due to the home run-friendly ballpark factors of Coors Field, I don’t think anybody denies that he is a great hitter. Like Bay, he is right-handed, hits for power, and takes walks. Throughout his career, Holliday has consistently hit for a better average, and is generally regarded as a better fielder than Bay. He will probably command a hefty asking price, very likely over the $100 million mark.

1. Roy Halladay

Halladay is the only player on this list that is not a free agent. Since last summer, the Blue Jays have been trying to trade him, and if a deal is going to get done, there’s a good chance that it will happen during the Meetings. The Red Sox and the Yankees are both in hot pursuit of Halladay, but it will take a lot for the Doc to be traded. The Daily News reported a few weeks ago that the Red Sox would need to part with Clay Buchholz and Class-A phenom Casey Kelley, while the Yankees would probably have to deal from a group including Joba Chamberlain, Phil Hughes, and prospects Austin Jackson and Jesus Montero. Both teams have recently shown an unwillingness to part with young talent, so it’ll certainly be interesting to see what will happen here. If the Jays really want to trade Halladay, they might have to lower their asking price.

Although I think these four will get the most attention this week, there are some other players that will draw significant interest. Ben Sheets, who missed all of last year with arm trouble, will surely be looked at by plenty of clubs due to his previous dominance. Rich Harden, who has only thrown at least 150 innings once in his career, will be looked at due to his serious strikeout numbers (10.9 K/9 in 2009) but will probably have to settle for a two-year deal because of his injury history.

Obviously, it’s not likely for us to see all of these players sign somewhere this week. But, the way these things seem to go, once one big contract has been handed out, the economic standard has been set, leading to more signings. Whatever happens, this week will surely have those avid baseball fans on the edge of their seats.

The Braves' new closer

With the likely departure of Type A free agents Mike Gonzalez and Rafael Soriano, the Braves needed to make a move to solidify their bullpen. Yesterday, they announced the signing of Billy Wagner to a one year contract with a club option for 2011. I think that it's a great move for the Braves, who will have their closer position set for at least a season. Although Wagner is 38 (turning 39 next July), it seems as if he has a lot left in the tank. The sample size from last year is extremely small (15 2/3 innings over 17 games), but I do think it shows something. Here's some PITCHf/x data from Wagner's 2009 stint with the Mets and Red Sox.

Billy Wagner Pitch Data and Results

	Average Speed (mph)	Max Speed	pfx_x	pfx_z	spin_dir
FF	94.11	97.3	9.45	8.01	130.77
SL	83.05	86.1	-2.33	-2.35	266.85
CH	85.1	87.5	9.16	6.14	124.25

Pitch	Pitch#	Pitch%	Swing%	Whiff%	Wide Zone%	Chase%	Watch%	SLGCON	RSv	RSv/100	xRSv	xRSv/100
FF	186	67.9%	43.5%	27.2%	53.2%	33.3%	45.5%	.636	1.52	0.82	1.60	0.86
SL	76	27.7%	46.1%	37.1%	28.9%	62.9%	40.9%	.000	2.82	3.71	1.27	1.67
CH	12	4.4%	25%	33.3%	16.7%	100%	100%	.000	0.00	0.03	0.00	0.03
	274	100%	43.4%	30.3%	44.9%	43.7%	45.5%	.500	4.34	1.58	2.87	1.05

Of course it's only 274 pitches, but it's extremely encouraging for someone who was one year removed from Tommy John surgery. The fastball velocity and extraordinary whiff rate show that he was still able to get it by hitters last year, and it's reasonable to assume that he will continue to gain arm strength as he recovers from the surgery. Of course, players at the age of 38 are always a risk, but I still think that paying $7 million for a pitcher that can be that dominant is a good deal.

Gameday PITCHf/x data is from MLB Advanced Media; it can be easily accessed via this tool.

Classifying relievers, part 2

Earlier this week, I did an introductory post on relievers and how I use AIPGR and Leverage Index to classify them. In this post, I’m going to look more closely at the members of the bullpen.

Defining relief roles

The closer of a relief staff is fairly easy to identify, since unlike other relievers, closers are judged by saves. So, by setting a baseline for a minimum number of saves, we can determine which pitchers are “closers” (for this post, I’ve set the minimum save number at 20). One-out-guys are also pretty easy to identify using AIPGR. Even though they’re often called that, these pitchers never actually wind up getting one out per appearance on average. Since 2005, only one qualified reliever has actually had an AIP under .5 (Mike Myers had an 0.49 on the 2006 Yankees). I decided I would set an arbitrary line at 0.80. I’ve always assumed that long-guys were used in low-leverage situations in which the only requirement was to eat up innings, and that primarily seems to be the case. Again, I've set an arbitrary AIP cut-off for defining long-relievers. I chose 1.40 as a reasonable number. Finally, we have our middle relievers, which don't fit any of the requirements of the other types of pitchers. I defined "middle relievers" as pitchers with AIP scores between 0.80 and 1.40 that do not meet the save requirement of closers. I split middle relievers into two categories, one with gmLI scores under 1.20, the other with scores at or over 1.20. Here are average numbers for qualified relievers since 2005 in each of the categories.

Role	AIP	pLI	inLI	gmLI	exLI	WPA	WPA+	WPA-	Adjusted WPA*
Closer	1.01	1.87	1.64	1.68	2.00	1.75	9.49	-7.75	1.22
Specialty	0.69	1.04	0.87	1.25	1.32	0.12	3.79	-3.67	1.02
Longman	1.58	0.75	0.67	0.90	0.91	0.17	3.90	-3.73	1.04
Setup1	1.02	1.37	1.19	1.41	1.46	0.87	7.24	-6.37	1.14
Setup2	1.06	0.88	0.78	0.97	1.08	0.04	4.26	-4.22	1.00

*WPA+ / |WPA-|. I use it as a rate stat in which 1 indicates a "neutral" contribution to the team.

Again, these are completely arbitrary cutoffs in terms of Leverage Index and AIPGR, but I think it gives us a good baseline nonetheless.

Are there any long relievers that actually pitch meaningful innings?

Long relievers have always been particularly interesting to me. While the numbers suggest pitchers with high AIPs are usually used in "mop-up" situations, there are some clear exceptions. Alfredo Aceves of the World Champion Yankees and J.P. Howell of the AL Champions Rays came to mind as examples from the past two seasons. I decided I'd look back at the numbers, set a minimum gmLI (I chose 1.20 because it seemed comparable to the setup and specialty relief scores), and see what came up. Here are the six pitchers that fit these criteria over the past five seasons:

Name	Year	AIP	pLI	gmLI	inLI	exLI	WPA	Adjusted WPA*
Travis Harper	2005	1.41	1.11	1.49	0.91	1.17	-1.02	0.86
Alfredo Aceves	2009	1.92	1.07	1.37	0.98	1.18	2.18	1.48
Zack Greinke	2007	1.40	1.07	1.37	0.99	1.40	2.14	1.90
J.P. Howell	2008	1.40	1.26	1.36	1.10	1.49	3.33	1.62
Pete Walker	2005	1.71	0.97	1.23	0.87	1.36	-0.10	0.98
Julio Mateo	2005	1.54	0.98	1.20	0.85	1.14	0.32	1.06

That's a pretty small number of pitchers, and of the six, only Greinke, Aceves, and Howell enjoyed much success.

One final, brief discussion. Sometimes I like to use average pitcher entrance inning (or whatever you want to call it), which can be calculated by averaging the game state in terms of innings for each appearance by a pitcher. I'll use Mariano Rivera as an example. In 2009, Mo entered most games with no outs in the 9th inning, which would mean that his average pitcher entrance inning for those games is 9. He had some games where he entered in the 8th, some others where he entered with outs in the 9th, and one where he entered in the 10th. All of this averages out to a 8.99 average pitcher entrance inning score, which is the mark of a clear closer. I didn't make any charts for this since I don't have a lot of data on it, but it is a pretty good way of further distinguishing bullpen pitchers.

Data is from Fangraphs and Baseball-Reference.