Here we go with Round 2. As a reminder, the Red Sox, based on the newly released PECOTA projections, are supposed to go 101-61 in 2008.
The Yankees, like the Red Sox, enter the season with few new names — just a lot of young ones getting older and more experienced. Also like the Red Sox, the Yankees were unlucky when it came to one-run games — which led to an underperformance of their Pythagorean record, and should be kept in mind when we get to projecting a record at the end of this post.
All that said, let’s get to it.
The Offense
First, the 2007 lineup (incidentally, I hope you yfs are happy. It pains me every time I click on the Yankees’ Baseball-Reference page). Determining the typical Yankee lineup for 2007 is difficult. Fourteen times, New York ran a lineup with Andy Phillips at first, Melky Cabrera in center and Johnny Damon at DH. But 12 times, Mientkiewicz was at first, Damon in center and Giambi at DH.
On the other hand, by batting order, Damon led off far more often than anyone else (123 games, to 32 for Cabrera), while the most common batting order featured Andy Phillips in the 8 spot even though he was supplanted by Giambi and Duncan by the end of the year. Nevertheless, I’ll go with the lineup trotted out there the most times in 2007 — and that is below:
|
Name |
2007 stats |
| Damon | .270/.351/.396, 12 HR, 41 XBH, 66 BB, 93 R |
| Jeter | .322/.388/.452, 12 HR, 55 XBH, 206 H, 15 SB |
| Abreu | .283/.369/.445, 16 HR, 61 XBH, 84 BB, 25 SB |
| Rodriguez | .314/.422/.645, 54 HR, 85 XBH, 95 BB, 156 RBI |
| Matsui | .285/.367/.488, 25 HR, 57 XBH, 73 BB, 73 K |
| Posada | .338/.426/.543, 20 HR, 63 XBH, 74 BB, 171 H |
| Cano | .306/.353/.488, 19 HR, 67 XBH, 39 BB, 301 TB |
| Phillips | .292/.338/.373, 2 HR, 10 XBH, 185 AB |
| Cabrera | .273/.327/.391, 8 HR, 40 XBH, 43 BB, 13 SB |
Formidable indeed. A better line for that eight spot might be the Yanks’ overall numbers at first base (which would duplicate some of Damon’s numbers): .284/.350/.419, 16 HR, 43 XBH, 86 RBI.
A lineup putting up those numbers should have produced 5.98 runs per game, or 969 total runs. The Yankees actually scored 968. So not bad at all. I would guess this is skewed a bit by using Phillips and leaving Giambi on the bench, but it shows the Yanks did not suffer much of a dropoff from their bench in 2007.
Let’s see how a prospective 2008 Yankee lineup would look:
|
Name |
2008 projections |
| Cabrera | .283/.342/.404, 10 HR, 40 XBH, 44 BB, 14 SB |
| Jeter | .297/.365/.407, 8 HR, 41 XBH, 169 H, 13 SB |
| Abreu | .276/.378/.435, 15 HR, 46 XBH, 81 BB, 18 SB |
| Rodriguez | .294/.401/.550, 36 HR, 72 XBH, 94 BB, 116 RBI |
| Matsui | .286/.367/.465, 18 HR, 48 XBH, 63 BB, 67 K |
| Posada | .287/.380/.479, 19 HR, 48 XBH, 66 BB, 129 H |
| Cano | .299/.340/.455, 15 HR, 54 XBH, 33 BB, 257 TB |
| Giambi | .235/.363/.453, 15 HR, 25 XBH, 46 BB, 300 PA |
| Damon | .278/.355/.417, 11 HR, 39 XBH, 55 BB, 15 SB |
Not nearly as formidable. The DH/1B situation is a mess, so I chose the old reliables of Damon and Giambi. Obviously, they’ll be heavily rested in favor of players like Duncan, Betemit, or maybe Ensberg. Still, with Giambi projected to a .796 OPS and Damon to a .772 OPS, it seems unlikely the backups will bring much more than that.
Otherwise, the players you’d expect to regress — Posada and A-Rod — do. It’s not incredibly optimistic on the young guys either, though. Cano and Cabrera are not slotted for any kind of significant improvement — in fact, Cano is slotted for a significant regression (for his age). Such a lineup should score 5.64 runs per game, or about 914 total runs. That’s 55 runs off, or between five and six wins.
To be fair, let’s also look at Dave Pinto’s calcs for the Marcel and Bill James projections, like we did for the Sox. Pinto inputs a lineup with Damon as the leadoff and Cabrera at No. 9. Since they’re projected to have similar numbers, I doubt it made much difference. With Giambi, Marcel projects a slight loss while Bill James sees things much rosier, giving the Yanks 1,004 runs. That’s between three and four wins better than 2007.
Averaging the three systems together, we get -12 runs — or about one win worse.
The Pitching
We all know the Yankees will field a solid, if not as intimdating, lineup in 2008. The question is whether they can improve on this performance:
|
Name |
2007 stats (ERA/WHIP/BAA) |
| Wang | 19-7, 3.70/1.29/.261, 199.1 IP, 104 K, 59 BB |
| Pettitte | 15-9, 4.05/1.43/.282, 215.1 IP, 141 K, 69 BB |
| Mussina | 11-10, 5.15/1.47/.305, 151 IP, 91 K, 35 BB |
| Clemens | 6-6, 4.18/1.31/.262, 99 IP, 68 K, 31 BB |
| Hughes | 5-3, 4.31/1.28/.238, 72.2 IP, 58 K, 29 BB |
| Igawa | 2-3, 6.25/1.67/.285, 67.2 IP, 53 K, 37 BB |
When only one starter in six with at least 10 starts manages an opponents’ batting average below .260, that’s not generally a good sign.
The top four bullpenners:
|
Name |
2007 stats (ERA/WHIP/BAA) |
| Rivera | 3-4, 3.15/1.12/.253, 71.1 IP, 74 K, 12 BB |
| Proctor | 2-5, 3.81/1.51/.245, 54.1 IP, 37 K, 29 BB |
| Vizcaino | 8-2, 4.30/1.45/.237, 75.1 IP, 62 K, 43 BB |
| Bruney | 3-2, 4.68/1.62/.238, 50 IP, 39 K, 37 BB |
The Yankees gave up 777 runs in about 1,451 innings. The weakness and injuries of the starters led to the overuse and ultimate failure of the bullpen. An off year by Mariano Rivera’s standards didn’t help things either. Roger Clemens’ return was a qualified failure (in the sense that he provided very little except to be better than Kei Igawa).
Now, the Yankees will rely heavily on Hughes, Ian Kennedy and Joba Chamberlain to keep them in the 2008 race. Does PECOTA think they can?
|
Name |
2007 stats (ERA/WHIP/BAA) |
| Wang | 11-9, 4.37/1.45/.272, 178.1 IP, 95 K, 59 BB |
| Pettitte | 12-8, 4.22/1.40/.262, 176 IP, 117 K, 60 BB |
| Hughes | 10-8, 4.42/1.39/.243, 152 IP, 129 K, 65 BB |
| Mussina | 9-8, 4.54/1.36/.268, 149.2 IP, 101 K, 40 BB |
| Kennedy | 9-7, 4.24/1.40/.234, 141 IP, 120 K, 68 BB |
| Chamberlain | 9-6, 3.71/1.24/.224, 145.2 IP, 162 K, 55 BB |
Much like the Red Sox with Clay Buchholz, the lesson from PECOTA is get Joba Chamberlain into the rotation ASAP. He projects to have better numbers across the board than any other pitcher in the Yankee rotation. Otherwise, Yankee fans would be thrilled to see "The Franchise" turn out to be the third-best of the three rookie pitchers. PECOTA essentially wipes the extremes and projects the veteran starters in the fair-to-middlin’ range.
And the ‘pen:
|
Name |
2007 stats (ERA/WHIP/BAA) |
| Rivera | 5-4, 2.69/1.13/.229, 60.2 IP, 51 K, 15 BB |
| Ramirez | 4-3, 3.61/1.31/.209, 60.2 IP, 76 K, 31 BB |
| Farnsworth | 3-2, 3.71/1.29/.232, 53 IP, 51 K, 21 BB |
| Veras | 3-3, 4.66/1.50/.253, 47.1 IP, 39 K, 24 BB |
The bullpen looks to be much improved, and PECOTA even has faith in Kyle Farnsworth.
Extrapolating those numbers over the 1,451 pitched by Yankee hurlers in 2007 gives the staff 668 runs allowed. That’s a lot better — better than every team in ’07 except Boston and 109 runs better than 2007. That’s an 11-win improvement, which intuitively seems unrealistic. And there’s a reason why.
In the Red Sox study, I extrapolated the top 10 Boston pitchers’ numbers over the total team innings to see if there was a gap between their expected runs allowed in that many innings and the actual team RA, and there was none. Doing the same with the Yankees, however, leads to a projected runs-allowed total of 692 — 85 fewer runs than the Yankees actually gave up in ’07. Just as the Red Sox benefited fom an unusually strong showing from call-ups and back-end relievers to essentially allow no drop-off from the Top 10 to the rest, the Yankees had plenty of Chase Wrights and Tyler Clippards and Jeff Karstens(es).
So, acknowledging that the Top 10 Yankee pitchers should have allowed only 692 runs (had they been able to pitch every inning) in 2007, the Top 10 Yankee pitchers should give up 668 — a gap of 24 runs, or between two and three wins. It’s safe to assume that the Yanks will not be turning over so many innings to such poor pitchers in 2008, so let’s take the average of 2.5 and 11, and say the Yankees will be seven wins better on the pitching side, based on PECOTA projections. It’s not terribly scientific, but it’s reasonable.
Conclusion
So let’s wrap this up with a handy chart for the Yankees.
| 2007 | 2008 | Diff | |
| RS | 969 | 957 | -12 |
| RA | 777 | 707 | -70 |
That’s a net gain of just under six wins for the Yankees from their 2007 Pythag record of 97-65.
The final standings, according to the PECOTA extrapolations:
Yankees 103-59
Red Sox 101-61
I’m not going to get into how likely this is, just from the standpoint of seeing two 100-win teams in the same division. Suffice it to say, it will be another dogfight in the AL East this year. Let the Great Debate begin!
YES!!!! Paul has projected the Yanks as the AL East winners!!!!
I kid. Thanks for all the work, Paul. I’m not as optimistic as PECOTA is–I think the Yanks will be good, but this is unbelievable.
A banner day for YFSF thanks to all of this amazing work from Paul.
Hah! Thanks for doing this Paul. I know it must have burned, but I’m glad to have your insights.
Thanks again so much for all your efforts, Paul.
Hate to point out an error, but for the 2008 predicted pitching numbers you wrote “2007” again.
Aside from that, amazing work. I’m somewhat skeptical too that there will be two 100-win teams in the same division, but it should be very exciting this year.
It’s also interesting that PECOTA predicts both IPK and Joba will have better years than Hughes. That’s something else I’m skeptical about.
I think the relative PECOTA projections for IPK, Hughes and Joba are very much age-related. Joba and IPK might be more major league ready–although, IPK and HUghes are essentially a wash–but I’d imagine the ceiling for Hughes and Joba is similar. Although, PECOTA is kind of bat shit crazy about Joba.
Wow nice work on this, I really enjoyed reading it. Thank you.
Yeah, Joba had an amazing run as a set-up man, but I’m genuinely surprised that PECOTA thinks he’ll have great success in his first year as a starter. I’m not trying to hate on Joba, I just think he will take longer to excell as a starter.
Another thing that surprises me is how few wins PECOTA projects for the starters. Honestly, 12 games is the most a starter will win? Impossible.
i’m projecting 8 days till pitchers and catchers report.
well – I was a little pissed when they predicted the huge drop off of Beckett’s numbers from 2007 to 2008, but I see they did the same thing to Wang as well, so I guess it is just how the system works. I was also pissed that they thought Manny would have the same type of HR year in 2008 as 2007, but then I see Arod’s numbers drop 18 – so again, the system is consistently shitty at least.
Oh, and btw, thanks for putting this (and everything else) together!
PECOTA — and all projection systems, for that matter — tend to be very conservative with wins. I put them there for consistency’s sake, but they’re generally pretty useless.
I’ll still be predicting the Red Sox to win the AL East in 2008, fwiw. I think PECOTA’s being too generous with the Yankee pitching and way too pessimistic with the Red Sox’ lineup.
I project the Yankees to break the MLB record for wins in a season and to win the AL East title handily. The Yankees will then proceed to sweep through the playoffs and win the World Series without losing a single game.
All will be right in the Yankee Universe.
This is not a reverse jinx–honest. ;-)
From Fire Joe Morgan:
http://www.youtube.com/watch?v=bHPlYnpiByA
Hilarious.
“…I’ll still be predicting the Red Sox to win the AL East in 2008,…”
bookmarked…
If the Sox and Yanks combine for 204 wins, there would likely have to be some ugly records in the rest of the A.L. East.
Agreed, Hudson. Which is why the Yankee projections are clearly too optimistic. ;-)
Paul, both of these were very nice reads.
OT: I’m sure there will be a McNamee’s Bloody Syringes thread *eventually* on this site…
The pitching and bullpen for the Yankees are indeed way too optimistic. However, it’s countered by the across-the-board 15 percentiles for each Yankee hitter. It looks to be just as pessimistic for the Yankee sluggers than the Sox sluggers. I’d definitely project the offense to be better, but the pitching, especially the bullpen, to be worse. If the pitching does not allow over 700 runs, it will be a miracle.
Well, considering the O’s will probably lose ~100 games this year, maybe 100 wins apiece for NYY and BOS isn’t so far-fetched…
Thanks for the great work, Paul!
yeah, the most volatile projections would probably be for young pitchers and middle relievers. This is very optimistic for the Yanks, and, unfortunately, I don’t buy it.
Sox REPEAT!
…nuff said!
Tampa bay finishes ahead of the Stanks
Heads roll and LA Joe (along with his fav Proctor) gets the last laugh!!!!
I am skeptical about:
Cano’s regression
ARod’s regression being so large
Hughes’ poor performance
Veras being the 4th best arm in the pen (I’ll be surprised if Ohlendorf is not among the top 4)
Wang’s poor performance
I think PECOTA is too optimistic for:
Kennedy
Ramirez
maybe Farnsworth
And that’s about it.
Maybe I’m just an optimistic Yankee fan, but I think things will be considerably better than PECOTA predicts. Maybe not better than 103 wins, but still.
Paul – I’m curious. You split the difference between Marcel, Pecota, and Bill James’ numbers for the Yankees offense, but you stayed with just Pecota numbers for the Yankees pitching. Maybe I’m not understanding clearly, but that seems inconsistent to me. What would the Yankees pitching numbers be if you split the difference between the three projections?
Hey MattyMatty,
My intent was to use PECOTA only on both pitching and offense for both teams because it worked so well last year. But as I went through the offenses, PECOTA looked unusually pessimistic on both teams’ projections, and I knew Pinto had already done some work with the other systems.
PECOTA does seem unusually rosy on pitching, especially with the Yanks, and I could do some averaging with other systems, but I’d have to do the work myself and, well, I’m lazy. :-P Maybe I’ll work on something this upcoming week, if I have the time.
It seems PECOTA is always pessimistic when it comes to hitting. I think that has to do with the historic norms that its working with. They tend to even out the bumps, so to speak, in players performance. You almost never see a Pecota projection that says a player who is older and had a down year (Drew, for example) has a career year. You might see the projection show a bit of improvement (say, due to BABIP or something), but mostly it seems that PECOTA is a pretty conservative (or some might say pessimistic) projector.
Thats essentially what I was trying to get at, that using one method for one team and then a different method for the other team strikes me as inconsistent.
The other thing that you didn’t take into account (for reasons you outlined yourself, i.e. you don’t want to spend 20+ years on this) is potential for injury. Off the top of my head I don’t think this favors either team, but it would be interesting to see how it plays out.
Plus, I wonder what Schilling’s injury does to the Sox. Are they better now (I know PECOTA loves Buchholz, as do I)?
Thanks for the hard work.