WS Coverage: Philadelphia’s Game 7 Pitcher
Game 7, of course, is always living in the realm of “if necessary,” so for all we know any and all analysis on the topic could be moot by tomorrow. However, let’s humor ourselves for a bit. With Cole Hamels having pitched Game 3 on Saturday, he would be on four days’ rest on Thursday for Game 7. With the Phillies now having committed to Pedro Martinez for Game 6, that leaves Charlie Manuel with the choice between J.A. Happ and Cole Hamels.
The real answer here is both. There is absolutely no reason, in a Game 7 situation with essentially every pitcher available besides the Game 6 starter, that a starting pitcher should be allowed to face the lineup more than one time. This is the reason that the league average FIP for starters (roughly 4.45) is so much higher than that for relievers (roughly 4.20). This difference is even more exaggerated when we consider short relievers and remove mop-up types from consideration. This is because the short reliever does not need to pace himself, and doesn’t have to face a lineup the second time, and especially not a third time, where the starter suffers even more.
Apart from Hamels (3.72 FIP as a starter) and Happ (4.46 FIP as a starter), the Phillies could also call on Joe Blanton (4.45 FIP as a starter) off three days rest to pitch an inning or two. They also have Chan Ho Park (2.10 reliever FIP), Ryan Madson (3.23), and Brad Lidge (5.45 FIP, but 3.19 career).
Between these six pitchers, the Phillies could easily get nine quality innings. The difference between starting and relieving comes out to nearly a run, but just for the sake of argument, let’s use a more conservative estimate of .75. Then we see Hamels at roughly 3.00 FIP, and Happ and Blanton sitting at roughly 3.75. Park is probably not a 2.10 FIP true talent, and probably is closer to his 3.90 average from the last two years. Madson has been at 3.20 the last two years, but was at 4.20 as recently as 2007. He fits closer to 3.50 FIP for our estimate. Lidge, on the other hand, has a 2.19 FIP season in 2008 and a 3.88 FIP season in 2007. Giving weight to recent performance, a good estimate for Lidge would be 3.95.
In that case, consider this pitching plan. Hamels starts, pitches three innings. Happ relieves for two innings, and then hands off to Blanton, Park, Lidge, and Madson in some order to finish the game. At no point do the Phillies have to rely on a player with a reliever FIP above 4.00. Leaving Hamels or Happ in past the first part of the lineup would certainly result in a worse pitcher on the mound in the 3rd or 4th inning.
Regardless of how they do it, the Philadelphia Phillies will need any competitive advantage they can get if they reach a Game 7, where they will be facing one of the league’s top pitchers in C.C. Sabathia, and will also have to deal with Mariano Rivera at some point in the game. Runs will likely be at a premium, and this is the Phillies best bet at suppressing the Yankees offense in Game 7.
Print This Post
I don’t understand why managers don’t manage like this more in elimination games. I’d even use Cliff Lee for an inning or two.
Just keep throwing fresher arms out there.
B-R pretty much takes care of this point for me:
Times facing opponent in game, 2009, MLB:
1st PA: 107,978 PA, .253/.328/.398, 2.08 K/BB
2nd PA: 44,477 PA, .270/.335/.434, 2.02 K/BB
3rd PA: 31,192 PA, .282/.347/.460, 1.83 K/BB
There may be some reliever selection bias in the 1st PA (.268/.332/.430 vs. starters against .252/.334/.395 vs. relievers), but it seems pretty clear that by at least the 3rd go around, it becomes much harder to record those outs. 12 points of OBP, 26 of SLG, and a 9.4% drop in K/BB is nothing to sneeze at.
I was going to bring up Cliff Lee, too. Why wouldn’t you use the best pitcher on the staff for an inning or two? It’s game 7 of the World Series, there’s no reason to hold anything back at that point.
Also, to further illustrate the points you made about time through the lineup – there’s actually going to be some survivor bias in those numbers I would imagine – higher quality starting pitchers should make up a higher proportion of the 2nd and 3rd time through the lineup, since they’ll pitch deeper into games more often….
Yeah, you can actually see that in the 4th+ time faced (which is why I didn’t include it, the only time a pitcher would see a guy 4 times in a game is if he’s dominant…or Dusty Baker is managing your team).
Cliff Lee should absolutely be used for an inning or 2. It would be his normal day to throw anyway, if in midseason, so they could get 30 pitches or so out of him, I’d think.
Great points. This also should allow for Manuel to pinch hit for the pitcher’s spot everytime through the order. Hamels career wOBA is 177.
You know what I hate, though? It would make sense for the Phillies to use this strategy tonight, too. But if they do, and they lose, it becomes on the manager according to the media, so no manager would do it.
As much as I criticize some managers, the lack of autonomy to make probabilistic strategy calls must suck. Instead, just do it the way it’s always been done to avoid the possibility for backlash.
If the single pitcher hits for the Phillies tonight or tomorrow, Charlie Manual should be fired.
In fact he should be thrown into the Bronx River.
Even if Pedro goes 1-2-3 in the first 2 innings with 5 SO and 0 BB?
I would probably let Pedro hit once tonight, unless he sucks on the mound. Matt Stairs hero time.
wait
fuck
I keep forgetting AL rules.
I’m an idiot.
Compare Cliff Lee’s numbers this year with some of the bench players on the Phillies. I’d let Cliff PH against Joba or Hughes over Bruntlett.
if Manuel allows a pitcher to hit either in game 6 or 7 he should be fired unless we’re in the 15th inning or something.
Good grief — like “Joe R” says, it’s an AL game. Of course the pitchers aren’t going to be batting. Why is everyone talking about firing Manuel or throwing him in a river when this is a complete non-issue to begin with?
WY, you’ve answered your own question. Now read everything again carefully.
I already did read it. It’s a silly argument to begin with.
WY, that’s why he ought to be fired. Relax!
So the Yankees should take out CC after the 3rd inning?
No, because he’s well above average.
Hamels/Happ’s pitching production can be replaced by a reliever on a short term run, and over a course of 9 innings, can be fully supplanted, you don’t need a starter to just eat 6-7 innings.
Sabathia’s pitching cannot be replaced via bullpen as easily, since he’s a superior pitcher. It *could* make sense to take him out in the 5th, and go with Robertson, Hughes, and Rivera from that point on, though.
phillies have more depth on that front, too. if you take out sabbathia you can’t replace him with pettitte or burnett(like the phillies could). the only bullpen guys that have shown any abillity for the yankees are rivera and marte. the phillies have beat up on the rest of their pen.
so its hamels/happ/blanton/lee/park/madson/eyre against sabbathia/rivera/marte/maybe joba?
advantage phillies. and if sabbathia gets knocked around early then they’re really in for it while the phillies have the arms to go deeper in the game if hamels gets hit early.
I was more on the front that Sabathia is more important to winning than Hamels and Happ are, and it’s harder to replace a top 10 MLB starter than an average one with a bullpen platoon, meaning the marginal gain of keeping fresh arms out there for the Yankees is immediately mitigated by losing Sabathias, while the Phillies see a much greater benefit since Hamels’ arm is much less important.
True, but the author says “There is absolutely no reason, in a Game 7 … that a starting pitcher should be allowed to face the lineup more than one time”. I think think Nick was responding to this blanket statement that covers both Hamels, Happ, Sabathia or any other Game 7 starter. I immediately had the same question too and searched to see if anyone else had commented on it.
vr, Xei
Just ran through the calculation, CC has a career 2.9 FIP (assuming about 4.11 PA/IP based on CC’s career average) the 3rd time through the order over the last 3 years. He actually does worse the 2nd time through the order, though not by much. So, obviously you don’t take out your 2.9 FIP pitcher for someone like Hughes unless you got a pretty good reason.
Maybe Hamels/Blanton/Happ rather than Happ/Blanton so you go L/R/L, then either Park or Myers rather than Lee to switch again.
That was supposed to have been Alston’s reason to start LHP Koufax on 2 days rest in ‘65 game 7 rather than RHP Drysdale on 3 day’s; that he would pitch once through, maybe pinch-hit and bring in the “real” starter with their ace reliever LHP Perranoski in reserve. But Koufax on the short leash instead pitched a 3-hit shutout.
Girardi would probably have at least Burnett thinking about throwing if there is a game 7, Sabathia is not automatic even though Hughes and Chamberlain should be capable of long relief. Even great pitchers on a roll should not expect to get 3 consecutive above-average performances against a playoff caliber team…Walter Johnson in 1925 would be an example, 1 run 11 H in 18 IP in games 1 and 4, 9 R 15 H 8 IP in game 7.
As far as “true” FIP, Madson somehow learned to throw a 95-97 mph fastball in 2008, so Madson in 2007 really isn’t that similar to 2009 Madson.
That’s actually really good managing.
Well not so much whoever was on Walter Johnson patrol.
This is way too out of the box for Manuel to consider, and he would get pasted by the media if it didn’t work. Sound concept for the Phillies though.
Like I said, it’s amazing how much influence the local media can have on day to day baseball operations.
No other sport does this happen. Mainly because in basketball and football, good or bad play and managing is clear and evident, in baseball, not so much.
I doubt they have much, if any, influence, especially on someone like Manuel.
When Francona temporarily moved Drew to the leadoff spot and Ellsbury lower, the media reaction ranged from apathetically amused by how things have changed (Bill Simmons) to piping mad. Batting order, which in the greater scheme of things isn’t all that big of a deal.
Joe R- You’re quite wrong about football, with the most obvious example being coaches’ aversion to going for it on 4th down, even though research shows they don’t attempt the conversion nearly often enough.
I think it would make more sense to do this tonight rather than tomorrow
There are two problems with doing it tonight. One is that if you do it and win, you might not be able to do it again tomorrow night. For instance, if you go Pedro / Happ / Blanton to get through 7 innings and use some combination of Madson, Eyre, Myers and Lidge to get through the 8th and 9th, you might not be able to use Happ or Blanton again tomorrow. The second issue is that the Yankees pitcher tonight is not as dominant as their pitcher tomorrow night, so you try to get 6 innings out of Pedro and save as much of your staff for tomorrow night if possible and necessary. Of course, since it is an elimination game, you do have to have a short leash on Pedro tonight if he begins to falter.
Yeah, if the Phillies are down early, everything is going out the window, or out on the mound if needed. Obviously you have to win 2 in row and have to manage understanding that if you win tomorrow there will be a game tomorrow, but if your chances of winning tonight start falling throughout the game, tomorrow’s game becomes less and less relevant.
whoops, that needs to be “if you win TODAY there will be a game tomorrow.”
“There is absolutely no reason, in a game 7 situation with essentially every pitcher available besides the game 6 starter, that a starting pitcher should be allowed to face the lineup more than one time.”
This is certainly an overstatement. There is at least one reason why the starter might be allowed to face the lineup more than once — namely, if that starter is really, really good and the pitchers in the bullpen aren’t. It’s that simple. Saying there is “absolutely no reason” is simply absurd.
There’s another reason not to simply take a guy out once you have gone throught the line-up once. And that is that the more pitchers you use in a game the more likely you are going to come up with an empty chamber. Not every pitcher is going to have it on a given day (I recall Bill James once wrote something to this effect, but he didn’t analyze data). For example, sure, Thursday is Lee’s day to throw anyway, but he is certainly not used to pitching in game situations on that throwing day. You won’t start him on 3 days rest but you’ll let him relieve on 2 days rest? When was the last time he relieved?
I am sure Cliff would do fine, and Philly could use the bullpen depth, but if Hamels is doing well, I would hesitate to take him out after he has faced 9 batters. And I would be very hesitant to go Hamels 9 batters, Happ 9, Blanton 9, Lee 9 (assuming the yanks get 9 or so baserunners at least), Eyre 1-2, Madson, etc.
This was my thought as well. I get the general argument, but the nagging counter to it is that one of these guys may simply have nothing that day. You always run that risk when you bring in a reliever. By committing yourself to bringing in several relievers FOR SURE, you are running the risk of finding that guy. I think limiting a pitcher to 1 time through the order is excessive. If he struggles, fine, absolutely pull him. Quick hook, definitely. But if he breezes through the order once, gosh, I can’t see why you would pull him. Burning through pitchers like Girardi on speed is a good way of seeking out the guy who sucks that day and putting him in a world series game.
As for using guys on their throw day, the Yankees actually did that with Pettitte in the regular season (either 2007 or 2008, can’t remember which). He pitched an inning or two in 1 or 2 games, and as I recall he pitched well.
But we can’t forget the possibility of extra innings. While there’s not a high probability, Manuel can’t risk leaving his team out to dry when all of his best pitchers have been spent and the dregs of his bullpen has to keep the Yankees offense off balance after the ninth. Using many pitchers for a few innings each is a great strategy, but Manuel’s got to keep one or two in store just in case, kind of like the All-Star game when the manager wants to get each pitcher in, but must keep one or two in reserve.
Speaking of future NY collapses…
Can someone point me to the game graph of 2001 Game 7?
Don’t think they have pre-2002 playoff data, so here’s the best I can do:
http://www.fangraphs.com/wins.aspx?date=2004-10-17&team=Red%20Sox&dh=0&season=2004
http://www.fangraphs.com/wins.aspx?date=2004-10-18&team=Red%20Sox&dh=0&season=2004
As a Yankee fan, I strongly support Mr. Moore’s plan to get Blanton, Happ and Lidge some innings in a potential game 7.
Particularly Happ. The Gods of BABIP Regression are ready to fix up some minor errors in J.A.’s 2009 line.