Raul Valdes for the rotation?

When Oliver Perez was removed from the rotation, Raul Valdes was mentioned as a possible replacement. In the end, the Mets went with R.A. Dickey, but Valdes came in for John Maine when he was removed for precautionary reasons, and pitched extremely well, striking out 6 in 5 innings. Is he a viable option for the rotation?

Stats:
Valdes, so far, has thrown 20.1 innings, hardly a big sample size. He has struck out 10.2/9 IP, and has a 3.3 K/BB ratio, both excellent figures.
As the above graph shows (click to make larger) Valdes gets excellent movement. That is one reason for his success, and his curve ball, with the most movement, is his outpitch, worth 2.8 runs.

The above graph again shows his movement, this time from the catchers perspective. The curve is a great pitch against lefties, though against righties, not as much, as it's down and in to them.

Conclusion
He has a somewhat worrying flyball trend that could regress, and I'd worried about him against righties. If he is going to start, the ideal conditions would be at Citi Field, with the large outfield, and against a lefty dominated lineup.

Could Jerry Manuel be making the right move?

Latest news from Port St Lucie is that Jerry Manuel is considering batting Jose Reyes third.
This is while Carlos Beltran would be injured

Manager Jerry Manuel said Friday that he might move Reyes to the third spot in the order from leadoff while outfielder Carlos Beltran is out after knee surgery. Either Angel Pagan, Beltran's likely replacement, or second baseman Luis Castillo would bat leadoff.

"I think he's ready for that if it's something we need," Manuel said, according to the Star-Ledger of Newark, N.J. "This could be a part of his evolution as a player."

"I'm good with it," Reyes said, according to the New York Post. "I just want to be in the lineup."

Reyes said the order switch isn't a definite yet, however.

"He's not sure yet. We just talked about it a little bit yesterday," Reyes said, according the New York Daily News. "I said I'm open to anything you want me to do. It's not a problem."

Could Jerry Manuel actually be making a good move in batting Jose Reyes third? In their book, The Book, Tom Tango, Andrew Dolphin and Mitchel Lichtman explored optimum lineup construction. They came up with this

Your three best hitters should bat somewhere in the #1, #2, and #4 slots. Your fourth-and fifth-best hitters should occupy the #3 and #5 slots. The #1 and #2 slows will have more walks than those in the #4 and #5 slots. From slot #6 through #9, put the players in descending order of quality.

CHONE has Reyes projected as the third best Mets hitter (excluding Beltran). So, in an optimal lineup construction, he should hit fourth. But how would Reyes hitting third look in a Mets lineup? I used Lineup Analysis to find out how many runs the Mets would score. The OBP and SLG columns used projections from CHONE
The Lineup was Pagan, Castillo, Reyes, Bay, Wright, Murphy, Francoeur, Catchers (.318 OBP, .371 SLG), Pitcher (.176 OBP and .179 SLG)
This Lineup would score 4.66 RPG. The optimum lineup, scoring 4.888 RPG would be Wright, Bay, Catchers, Pagan, Reyes, Francoeur, Murphy, Pitcher, Castillo.

If you switch Castillo and Pagan, having Castillo lead off, you score 4.7 RPG; switch Castillo and Pagan, and Bay and Wright, you get 4.8 RPG. Switch Bay and Wright, you get 4.65 RPG.

And if Reyes doesn't bat third, and the lineup is Reyes, Castillo, Wright, Bay, Murphy, Pagan, Francoeur, Catchers, Pitchers they'll score 4.66 RPG.

So if you're going to bat Reyes third, have Pagan lead off and Wright hit fifth, though it doesn't make that much difference if Wright bats fourth or fifth.
In the end, batting Reyes third or first is not going to make a difference to the Mets.

Jason Bay xBABIP projections

Continuing with the series, I used the xBABIP calculator to normalize the luck for Jason Bay, and to determine what his offensive output will be using this xBABIP and Chone projections.

Jason has a .352 xBABIP over his career, which is 20 points higher than his average BABIP. If we use his .352 xBABIP, Jason will bat .297/.402/.550 (AVG/OBP/SLG) which is a great year. He will hit 34 homers, and have a .410 wOBA. He is projected to have 622 plate appearences. If we use his average BABIP, .332, he'll bat .283/.391/.535, with a .399 wOBA.
Using his CHONE projections (.267/.376/.514, .388 wOBA, .307 BABIP) on Fangraphs, he is prjected for 4 WAR.
Using his xBABIP calculations and an average of his defensive projections from CHONE and Jeff Zimmerman's projections at Beyond the Boxscore, Jason has a total of 50.2 RAR, which leads to a 5 WAR. I think that is a little too optimistic, and his projected total from Fangraphs is probably right, 4 WAR.
CHONE, xBABIP normalized and BABIP normalized projections here.

Carlos Beltran season projection

Again, using the xBABIP calculator, Carlos' xBABIP is an astounding .428, which is way above his career norm of .307, and above his Chone projected .295 BABIP. I am using Chone projections, by the way, because it has been proven that Chone projects better than any other projection.
Back to Beltran: I'm going to calculate 2 adjusted season lines for Beltran: One with his .428 xBABIP, and one with his .307 BABIP.
Beltran is predicted to have 118 Games, 454 AB's, 126 hits, 27 doubles, 2 triples, 21 HR's, 79 RBI's, and have a line of .278/.362/.485. The totals are low because it's based on previous years to find an average, and last year Beltran missed a lot of games. Adjusted to a .428 BABIP, his line becomes .335/.412/.559. His wOBA is .419, which leads to 40 wRAA. He has a projected 1 UZR, gets a 1.8 positional adjustment and gets 17 Replacement runs. This adds up to 59.8 Runs, which divided by 10 gives Carlos a 6 WAR season, which is amazing.
However if his BABIP is .307, he bats .286/.370/.496 with a .378 wOBA. This leads to a 21.5 wRAA, which changes his WAR to 4 WAR, which is good, but not 6 WAR great.
However, it is still extremely productive, and I think his line will be something like .300/.390/.520. Remember, 2011 is the end of his contract, and personally, if Beltran is healthy he should be resigned.

2010 Mets Projections, Right Field

Sorry I haven't been writing recently, some stuff came up. This is a new series, where I will give some projections on the Mets for 2010, based on the CHONE projections. Today I am dealing with Jeff Francoeur.
Francoeur is projected to bat .270/.311/.421, in 596 AB's with 17 HR's and 84 RBI's . Because a lot all of his skill comes from his contact ability, BABIP is very important for him. So I looked at THT's xBABIP calculator, and based over his career, Francoeur should have a .310 BABIP. This is pretty close to his career of .307, and close to his projected chone BABIP of .300 (unfortunately, CHONE doesn't project Sac flies)
Seeing as his xBABIP is .310, that can change his slash stats, as he would now have about 166 hits instead of 161. An extra 5 hits would translate to (based on career norms) 4 singles and 1 double. That would change his slash stats to .297/.319/.431. His numbers lead to a .318 wOBA, which based on a .330 league average wOBA is -6.7 wRAA (to know what these mean, go here.) Then added to Beyond the Boxscore's projected UZR (fielding metric) of 2 runs, positional adjustement of -7.04 runs and his replacment level of 21.13 runs, we get a total of 9.39 Runs, which on the Runs to wins scale (10 Runs per win) leads to a .9 WAR. For more information on how this was computed, go to fangraphs.com, click on glossary and scroll down to wins above replacement.
.9 WAR, which on a $4.5 Million dollar per win scale means Francoeur is worth $4 million, which is about how much he's getting paid.
I'm all for giving Frenchy a chance, but if his BABIP is under .300, he's gonna suck majorly. I just hope Omar doesn't give him 3/36, like he did to Oliver Perez.
Next Up: Carlos Beltran.

trRA for hitters

Ever since Voros McCracken's ground breaking work on DIPS, and luck, there has been debate on how much control pitchers and hitters have on luck. BABIP measures whether a player has been lucky or unlucky, and there are many defense independent stats. One of the better ones is tRA or true Run Average. As this article shows, it is based on batted ball data. This same idea can be applied to hitters, and there can even be a regression value. The formula for this stat would be like wOBA; number of batted balls*batted ball run value and add this to each hitter and then divide by plate appearences. Walks and strikeouts can also be added. Using the run values given in the article about tRA, I applied this to David Wright and the Majors. For David Wright it comes out to .126; some work might be required on this statistical measure, but personally I think it is a good way to evaluate hitters, and rewards them for what they hit, rather than the outcome.

What the 2009 Mets could've been

Seeing as Omar and Jerry are staying, I wanted another good reason to have them fired. So, I decided to project the Mets win total in 2009 if no one was injured. I used Wins Above Replacement, which can be found at www.fangraphs.com. There were a couple of rules: Players who had had injuries last year that forced them miss significant time (10 games+ in one stretch) could get injured this year. Thus, Carlos Delgado, John Maine, Oliver Perez and JJ Putz all get injured in this year's version along with Ryan Church and Angel Pagan. Also, if a player was called up due to injury, and would not get called up unless in September does not appear. (Sucks for you, Argenis Reyes). Also, I am going by 4.5 PA/G for starters. wOBA stays the same for hitters, as does UZR.
For pitchers, I used the guide that fangraphs has. So, a pitcher's FIP (fielding Independent Pitching) is divided by .92 to convert it to a scale with Run Average. It is then divided by the Park Factor to normalize it (I think...). I used ESPN's Park Factors, which was .94 for Citi Field. Also, for the league RA FIP, I used 4.53, the NL RA.
So, the 2009 New York Mets, who do not lose Jose Reyes, Carlos Beltran, David Wright and Johan Santana finish 82-80. Feeling proud, Omar and Jerry?
Carlos Beltran has an awesome year, with a 6.5 WAR. David Wright has a 4 WAR. Because Fernando Nieve doesn't start, he gets 15 appearences out of the pen, 30 innings (2 per G) and has a war of 3.5. I know, it's fantasy, but I was just going by what I got, so bear with me.
Billy Wagner does not get traded for Chris Carter, and appears in 15 games. Ramon Castro still gets traded, Angel Pagan doesn't play in as many games, and Cory Sullivan doesn't play at all. Johan Santana and Mike Pelfrey each have 33 starts, Tim Redding gets 20, Nelson Figueroa gets 19 starts.
I know this isn't the most accurate way of figuring out what the Mets could've been, but it wasn't extremely hard, and still paints the main picture: The Mets are a .500 team, and Omar Minaya did not give this team enough starting pitching depth, and the starting pitching wasn't good enough. It also shows that this team needs a good offensive outfielder and first baseman, as we knew about. What I'm mainly trying to say here is that the Mets bad season this year was not just due to injury; it was also due to lack of good starting pitching, starting pitching depth, and the need for an outfielder who is good offensively.
Update: The spreadsheet

More on Murphy

I was looking at Daniel Murphy's stats, and noticed how he has a BABIP 100 points lower than last year's (.386-.286). So, using this (password is tuftsbat if you use it), I attempted to find out where Murphy's babip should be. It is expected to be .298 for this year, so he really should have about 117 hits for a .274 batting average. This would also increase his OBP to .324. To determine his slugging is a little more difficult. To do his, I have to figure out the percentage of hits that are doubles and triples. 26% of his hits are doubles, and 2.7% of his hits are triples. (Because a home run is not ball in play on a playing field, there is no increase in homers). So, if I round, 2 of his extra 6 hits are doubles, and there are no extra triples. So 4 hits are singles, and 2 are doubles. This gives him approximately 181 total bases, which gives him a slugging percentage of .423.
Murphy also had a lucky year last year, where he had a BABIP of .386. His xBABIP was .340, so he should've had 34 hits. That would've meant his average was .260, with a OBP of .351. 22% of his hits were doubles, and 7.3% of his hits were triples. That means he would've had 8 doubles, 3 triples, 2 home runs and 21 singles. That is a slugging percentage of .412.
Daniel Murphy's actual stats:

Year	PA	AB	1b	2b	3b	HR	bb	k	hbp	tb
2008	151	131	27	9	3	2	18	28	1	62
2009	473	428	70	29	3	9	35	62	0	173

avg	obp	slg	ops	Ops+	wOBA
.313	.393	.473	.871	129	.373
.259	.311	.404	.716	87	.309

OPS + is a stat that is adjusted for park and league. 100 is average, above 100 is above average, below 100 is below average (129 is 29% better than average)

Daniel Murphy's xBABIP stats

Year	PA	AB	1b	2b	3b	HR	bb	k	hbp	tb
2008	151	131	21	8	3	2	18	28	1	54
2009	473	428	74	31	3	9	35	62	0	181

avg	obp	slg		ops	Ops+	wOBA
.260	.351	.412		.763	112	.345
.274	.324	.423		.747	108	.323

Note: OPS+ used ESPN Park factors for Shea Stdium (.946) and Citi Field (.941)

All in all Daniel Murphy would be a perfectly adequate bench player/utility guy, if he can play second base. A Mark DeRosa type.

Mets chances at winning last 30 games

So to see what the Mets chances of winning the next 30 games are, I used a stat made by Bill James called log-5, and the strength of schedule concept from football. From the Mets next 30 games (against the Rockies, Cubs, Braves, Nats, Marlins, Astros and Phillies) the total wins and losses for those teams are 4630-460. That works out to a .502 winning percentage. As of today the Mets have a .447 winning percentage. The log5 method is:

             A - A * B
 WPct = ----------------- where A is Team a's winning percentage and B is team B's winning percentage.

        A + B - 2 * A * B

The Mets will be team A, and the opponents team B. This comes out to .445 or 44.5%. That means we have a 44.5% chance of winning those games. That means over the last 30 games we would go 13-17. Which is pretty good considering our team, and we would be 72-90. However, I don't think we will, seeing as the Opponents winning percentage is brought down by the Nationals horrible record, and our's is brought up by our good record at the beginning of the year.
Still, 13 wins will see us to 72, and I guess thats the goal.

Mets record

To find this out, I'm going to use a stat called WAR or Wins Above Replacement. It is explained here. A team full of replacement level players would have a .300 Winning Percentage.
The Mets batting and Fielding WAR is 7.7. The pitching WAR is 7.9. After 133 games (how many the Mets have played) a .300 WP would mean a team had a record of 40-93. (rounded). So the Mets have 15.6 Wins Above Replacement. Added to 40 gives the Mets 56 wins, which is pretty damn close to what they have (59). So, I'm not saying they should have 56 wins, I'm saying that is what the talent level on this team is.

Who has been the best since Johan went down?

Using FIP (fielding independent pitching, it is on an ERA scale), the best has been Nelson Figueroa, and the worst has been Ollie, and Parnell. Note, these are from all starts following Santana's last start.
Here is the link.

This is Painful. Lose 10-1

The big news of the evening was that Steven Strasburg signed a 4 year deal worth $15 million. Great. The whole point to lose these games was so we could get Strasburg. We were the only team that could afford him. The Mets signed there top pick, Steven Matz. Glad Fred Wilpon can still shell out $895,000 to him. You know what the Mets could get for $895,000? A pitcher who is better than Livan Hernandez (they could get someone like that for $400,000).
But, back to the game. Livan was terrible. Absolutely godawful suckyness. 5.1 innings of 6 run baseball. JERRY EXPLAIN TO ME WHY HE IS STILL ON THE TEAM?????????? He has a 5.72 DERA (Baseball Prospectus ERA scaled stat, defence independent, and I'm using it because I get minor league translations). You know who has a lower translated DERA for Buffalo? (And who is younger) Tobi Stoner (5.29), Kyle Snyder (5.29). You know who has that at Binghamton? Ryan Coultas (4.06) and Chris Mason (3.38). Not to mention Nelson Figueroa who is in the bullpen (gave up a run today). At this point Tim Redding could be better. Actually, probably not. Seeing as he gave up 3 runs in an inning. And we scored 1 run. And guess who is pitching tonight: Oliver Perez. When does it end? I mean, come on. I think when Livan pitches and when we have expanded rosters, the AAA AA team should play. Alex Cora went on the DL to have surgery for his ligament. Well, he showed grission by playing for 3 months with that injury.
Stupid Decision of the Day: Cory Sullivan getting thrown out at home in the first. Using Win Expectancy, the break even point is 124%. That means that he has to be safe on those plays 124% of the time for it to be a good decision. Obviously, thats impossible. So using Run Expectancy we see that the break even point is 77%. So Sullivan has to take the extra base successfully 77% of the time. And Francoeur was on deck. Francoeur probably would not have hit into a double play, as he's only hit into 1 as a Met. So Razor Shines, shame on you.
Random Thought of the Day: You know its a sad state of affairs when the lead story on Mets.com is about David Wright's leadership. When he is on the DL (he's the leader of the people on the DL, there's almost enough for a whole team!)
Clutch Play of the Game: Sheffield RBI single in the third. +.112 WPA(*100 to get%)
The Next Torture Session for Fans: Braves at Mets 7:10 PM ET
Oliver Perez (2-3 5.97 ERA, 5.83 FIP, 6.48 tRA) looks to make this torture quick.
Derek Lowe (12-7, 4.08 ERA, 3.77 FIP, 4.32 tRA) will throw a no hitter.

tRA now on fangraphs.com

I used to get it from statcorner.com, and still might, I'm not sure. The tRA on fangraphs will be a little different then on statcorner, because they use different batted ball info, and different park factors.
By the way, tRA is not on a ERA scale. To get it to a ERA scale (tERA) multiply the tRA by .92.
Here is some info on tRA. (Read the comments; some of them have more info)

Mike Pelfrey 2008 and 2009

After Last night's performance, Mike Pelfrey's ERA went to 4.88. This is a far cry from his ERA of 3.72 ERA last year. However, I think this has to do with the defense that has been behind him. His FIP (fielding independent pitching on ERA scale) was 3.96; this year it is 4.12, so there is not a lot of difference. This year, his K/9 is higher, his BB/9 is higher, but his groundball percentage is a percent higher, at 50.9 compared to 49.6. His HR/9 is a touch higher. This should account for the higher FIP, and the more groundballs might account for the higher ERA, seeing as the Mets defense has been pretty crappy this year. Another stat to measure pitchers is xFIP, and is available on the Hardball Times. Instead of HR, xFIP is expected FIP. Here is an article on the two. The formula for FIP is (13*HR+(BB+HBP)*3-K*2)/IP. Because pitchers sometimes have less control on home runs (due to park factors, altitude, etc) The Hardball Times replaced HR with the league average of HR/FB%, and used that to multiply flyballs. So the formula for xFIP= ((FB*HR/FB%)*13+(BB+HBP-IBB)*3-K*2)/IP. This is another good article. So Pelfrey's xFIP is 4.50, compared to 4.70 for last year. His BABIP is also higher this year, indicating he's been unlucky. So don't worry about Pelf, he'll be fine. ZiPS (Fangraphs.com projections) predicts that he'll have a 4.31 ERA for the rest of the year. So Pelf is fine, and is a good number 3 starter. Just not the number 2 guy.

Other Options to Close

After Frankie's recent meltdowns, I have decided to write a post exploring who would be a better closer. I'm using a stat called clutch, from fangraphs. It measures how well you do compared to normal situations, which is why it's called clutch. Seeing how being a closer means you are clutch, I'm going to see how clutch some of our bullpen members have been. The formula is WPA/pLI- WPA/LI. What the hell does that mean? WPA is win probability added. pLI is a player's average leverage. WPA/LI is the win probability divided by Leverage for every event in a game. Okay, still with me? The most clutch pitcher on the Mets is Johan Santana. Brian Stokes is the most clutch revliever, with a Clutch of .69. (The leader this year in the Majors, Arroyo (!!!) is 1.41)
Pedro Feliciano could be an option, with a rating of .26. K-Rod, btw is -.50. With such greats such as Takahashi, and Switzer.

A new stat I'll be using

Okay, I know this stat has been invented and used before, but because I can't find it, I'm going to do my own one. The stat is Predicted Winning Percentage, or PrWP. (Prip is fine to pronounce) and it takes the formula of Offensive Winning Percentage. The formula is: Team Runs Per game divided by DERA (look for the glossary) squared divided by team runs per game divided by DERA squared plus one: (TmRPG/DERA)^2/((tmrpg/DERA)^2+1). Their is also Predicted Winning Percentage above Average, or PrWPA, with average being 4.50 DERA, and has a neutral adjusted league average RPG. There is also Park and league adjusted PrWP, which is the same formula as above, except TmRPG, which is park adjusted, and DERA, which is park adjusted. Park adjustements can be found on ESPN.com, right here. 1 is a neutral park, above one favors hitters, and below favors pitchers. As you can see, Citi Field depresses run scoring a lot, ranking 26th, and is a severe pitcher's park.
An Example:
Johan Santana, 3.63 DERA. 4.26 RPG for the Mets
(4.26/3.63)^2/(4.26/3.63)^2+1)=.579*21 (his starts)=12 expected wins and 9 expected losses, which is quite close to his actual record of 12-8.

A little explanation

Okay, some of the terms I use are a bit hard to understand. So, I'm going to explain them.
OBP: On base percentage. Read Moneyball. It counts walks and HBP. League Average is .330, good is .350, very good is .360 +, and great is .400. H+BB+HBP/AB+BB+HBP+SF
SLG: Slugging percentage. TB/AB or 1b+2*2b+3*3b+4*hr/ab
OPS: on base plus slugging. OBP+SLG. Correlates the most with scoring runs.
EqA: Equivalent Average. A stat developed by Clay Davenport of Baseball Propsectus, it is like OPS, except on a batting average scale.
EqR: Equivalent Runs. The run companion stat to EqA. Think of it like RBI.
Runs Created: Created by Bill James, the simple form is OBP*TB or OBP*SLG*PA where TB is total bases and PA is plate appearences. Think of it like RBI in the amount.
FIP: Fielding Independent Pitching, (13*hr+3*bb-2*k)/ip+3.2. On a ERA scale, measures what the ERA should be, defense and luck independent.
tRA: trure run average. Maybe the best way to measure pitchers. It is hard to explain here, but it basically measures what the pitchers ERA should be based on the amount of times a pitcher gives up a groundball, flyball, strikes someone out, gives up a line drive, home run, or walks someone. Each event is give a run value and out value. More here.