Updated Yankee and Blue Jay Marcels Projections

I ran Marcels projections for the 2013 Yankees and Blue Jays, based on their current depth charts, and the results correspond to what most of us no doubt perceive regarding the offseason: the Yankees have gotten a bit worse while the Blue Jays have significantly improved. 

The Blue Jays are projected to have a 91-71 record, the Yankees 90-72.

Before proceeding with the numbers, let's put these projections in context. These projections do not say the Yankees "are an 90-win team" in a precise sense. Marcels is a good projection system with a proven level of accuracy within a reasonable range. However, the projections are not prophecies. Rather, they are probabilistic estimates of future performance. This makes them a useful tool - a mechanism that gives us some insight into the quality of teams' rosters - as we play armchair GM.


First, here are the team results. The Blue Jays project to outhit the Yankees, while the Yankees project to have better pitching:

Team	W	L	RS	RA
Blue Jays	91	71	784	691
Yankees	90	72	758	678

Runs scored were calculated using batting runs, and runs allowed using the pitchers' projected runs allowed. Expected wins-losses were then calculated using the Pythagorean expectation formula.

To get to these team numbers, I started with the projected stats for players on each team most likely to play significant roles. I then adjusted each team's totals to a full-seasons worth of plate appearances and innings pitched; the details are at the end of this post.

Following are the player projections. I selected the players using my knowledge of the teams plus their depth charts.

Yankee pitchers:
 

Pitcher	IP	G	GS	H	R	ER	HR	BB	SO	ERA	FIP	WHIP	HR/9	SO/9	SO/BB	Reliability
Aardsma, David	26	26	0	24	13	12	3	10	22	4.15	4.13	1.31	1.0	7.6	2.2	15%
Chamberlain, Joba	38	38	0	37	18	17	5	12	35	4.03	4.06	1.29	1.2	8.3	2.9	40%
Eppley, Cody	49	61	0	48	23	22	5	17	39	4.04	3.88	1.33	0.9	7.2	2.3	36%
Hughes, Philip	161	30	28	160	84	78	25	46	136	4.36	4.43	1.28	1.4	7.6	3	76%
Kuroda, Hiroki	190	29	29	186	82	75	23	49	147	3.55	4.00	1.24	1.1	7.0	3	81%
Logan, Boone	57	84	0	53	26	24	6	22	58	3.79	3.61	1.32	0.9	9.2	2.6	51%
Nova, Ivan	161	27	26	166	83	77	21	53	131	4.3	4.30	1.36	1.2	7.3	2.5	76%
Pettitte, Andy	98	16	16	94	43	39	12	31	82	3.58	4.07	1.28	1.1	7.5	2.6	55%
Phelps, David	87	29	10	76	37	35	11	31	79	3.62	4.13	1.23	1.1	8.2	2.6	51%
Pineda, Michael	77	13	13	65	35	33	9	24	77	3.86	3.66	1.16	1.1	9.0	3.1	54%
Rapada, Clay	46	85	0	42	22	20	5	18	41	3.91	3.97	1.3	1	8	2.3	35%
Rivera, Mariano	35	37	0	32	14	13	3	9	30	3.34	3.34	1.17	0.8	7.7	3.3	40%
Robertson, David	62	66	0	53	22	21	5	24	73	3.05	3.02	1.24	0.7	10.6	3.0	57%
Sabathia, C.C.	184	26	26	174	78	69	19	49	175	3.38	3.46	1.21	0.9	8.6	3.6	82%

Yankee hitters -  Since we know A-Rod will likely miss at least half the season, I halved the PAs Marcels projected for A-Rod, and prorated his other stats:

Hitter	PA	AB	H	2B	3B	HR	BB	SB	CS	BA	OBP	SLG	wOBA	BRs	Reliability
Cano, Robinson	617	560	169	39	3	27	46	5	2	.302	.360	.527	.369	100	87%
Cervelli, Francisco	215	189	50	8	1	4	19	3	1	.265	.338	.381	.312	24	56%
Diaz, Matt	286	262	65	15	1	5	18	4	1	.248	.302	.370	.290	29	67%
Canzler, Russ	249	227	60	10	1	7	18	3	1	.264	.319	.410	.311	29	30%
Gardner, Brett	277	240	64	10	3	4	30	19	5	.267	.354	.383	.329	35	78%
Granderson, Curtis	611	533	130	20	6	34	65	14	4	.244	.330	.495	.352	88	87%
Jeter, Derek	631	576	163	26	2	11	44	12	4	.283	.338	.392	.319	74	87%
Nix, Jayson	316	284	65	15	1	9	22	6	3	.229	.293	.384	.289	32	69%
Nunez, Eduardo	284	258	70	13	2	6	20	16	4	.271	.323	.407	.319	35	63%
Rodriguez, Alex	254	225	59	11	1	10	24	4	1	.262	.339	.453	.342	35	84%
Suzuki, Ichiro	604	567	155	23	4	7	30	28	6	.273	.311	.365	.293	64	87%
Stewart, Chris	297	267	63	14	1	6	23	3	1	.236	.299	.363	.284	29	56%
Teixeira, Mark	530	459	114	24	1	26	56	3	1	.248	.336	.475	.346	74	86%
Youkilis, Kevin	506	435	110	23	3	18	55	3	1	.253	.350	.444	.343	69	83%

Blue Jay pitchers: The sums of projected games slightly exceeded 162, I removed a few projected starts from their spot starters and correspondingly incremented their games relieved.

Pitcher	IP	G	GS	H	R	ER	HR	BB	SO	ERA	FIP	WHIP	HR/9	SO/9	SO/BB	Reliability
Dickey, R.A.	197	30	29	184	80	73	22	51	157	3.34	3.89	1.19	1	7.2	3.1	81%
Johnson, Josh	162	26	26	147	67	63	13	52	142	3.50	3.42	1.23	0.7	7.9	2.7	75%
Buehrle, Mark	182	28	28	190	86	80	23	43	112	3.96	4.30	1.28	1.1	5.5	2.6	81%
Morrow, Brandon	140	24	24	123	67	62	15	50	143	3.99	3.70	1.24	1	9.2	2.9	75%
Romero, Ricky	173	27	27	163	90	82	20	75	137	4.27	4.51	1.38	1	7.1	1.8	81%
Happ, J.A.	145	27	25	145	83	78	20	61	130	4.84	4.45	1.42	1.2	8.1	2.1	75%
Janssen, Casey	62	61	0	55	24	23	7	17	58	3.34	3.68	1.16	1	8.4	3.4	55%
Santos, Sergio	34	34	0	31	16	15	4	14	36	3.97	3.82	1.32	1.1	9.5	2.6	41%
Oliver, Darren	58	67	0	54	23	21	5	17	50	3.26	3.55	1.22	0.8	7.8	2.9	53%
Delabar, Steve	59	54	0	50	27	26	9	22	63	3.97	4.24	1.22	1.4	9.6	2.9	42%
Lincoln, Brad	81	50	2	80	41	39	12	25	68	4.33	4.38	1.30	1.3	7.6	2.7	59%
Rogers, Esmil	78	51	1	86	47	45	10	31	71	5.19	4.24	1.50	1.2	8.2	2.3	65%
Perez, Luis	54	41	0	53	27	25	6	20	47	4.17	4.09	1.35	1	7.8	2.4	48%

Blue Hay Hitters - The Marcels projections for the 2 catchers are based on both being regulars; so to make this realistic, I reduced Josh Thole's projected PAs by 40%, and prorated his other stats:

Hitter	PA	AB	H	2B	3B	HR	BB	SB	CS	BA	OBP	SLG	wOBA	BRs	Reliability
Arencibia, J.P.	435	399	94	19	2	20	29	3	1	.236	.290	.444	.311	51	77%
Bautista, Jose	465	387	102	19	1	31	70	7	2	.264	.381	.558	.391	82	85%
Bonifacio, Emilio	401	357	99	13	5	4	35	26	4	.277	.341	.375	.321	48	79%
Cabrera, Melky	521	478	145	28	6	12	35	12	5	.303	.349	.462	.343	73	85%
Davis, Rajai	477	442	114	24	4	7	27	39	10	.258	.305	.378	.306	54	82%
Encarnacion, Edwin	575	507	136	27	0	29	57	9	2	.268	.346	.493	.356	86	84%
Gose, Anthony	295	261	66	13	3	5	26	14	3	.253	.326	.383	.313	34	44%
Izturis, Maicer	409	371	96	21	1	5	30	12	4	.259	.319	.361	.300	43	78%
Lawrie, Brett	485	439	124	24	4	14	36	13	6	.282	.342	.451	.340	65	74%
Lind, Adam	431	396	98	17	2	17	29	2	1	.247	.299	.429	.310	50	83%
Rasmus, Colby	565	505	123	24	5	20	51	7	4	.244	.314	.430	.315	68	85%
Reyes, Jose	617	563	165	31	11	11	46	33	8	.293	.345	.446	.340	86	87%
Thole, Josh	250	223	58	11	1	3	22	1	1	.260	.331	.359	.294	26	77%

I'll be doing a similar exercise for the other AL East teams at some point.



Following are details explaining how projected team stats were estimated.

For hitting, I summed up the batting runs and plate appearances of the players listed above. The teams at this point less projected PAs than would occur in a season, so using batting runs for Pythagorean purposes would have been misleading. However, since the teams will utilize other "fill-in" players, and pitchers will get a few PAs, I adjusted the team stats to a full-season basis by doing the following:

• Calculated batting stats of 2012 AL pitchers.
• In 2012, AL teams averaged 23 pitcher PAs, so I prorated the league average pitchers' stats to 23 PAs.
• Added 23 PAs and the corresponding prorated stats to the sums of the stats of the projected players.
• In 2012, AL teams averaged 6151 PAs. So to arrive at a full-season's number of PAs, I subtracted the sum of the projected players' PAs from 6151, then subtracted 23 to account for pitchers. The remainder is an estimated number of PAs for fill-in players.
• I calculated the batting stats of 2012 non-pitchers with less than 300 PAs, then prorated those stats to the team's estimated fill-in PAs.

The following table shows how this worked:

Category	PAs	BA	OBP	SLG	Batting Runs
Blue Jays
Projected Players	5926	.267	.331	.434	766
Pitchers	23	.100	.143	.100	-1
Fill-ins	202	.230	.294	.350	19
Total	6151	.265	.329	.430	784
Yankees
Projected Players	5677	.263	.330	.427	717
Pitchers	23	.100	.143	.100	-1
Fill-ins	451	.233	.295	.355	43
Total	6151	.260	.321	.421	758

It's worth noting that the Blue Jays have a somewhat more settled projected roster. This is primarily because the Yankees currently have open spots for Matt Diaz and Russ Canzler, neither of whom have played enough to project as regulars, and also because A-Rod and Gardner have a low projected number of PAs due to their injury problems. This projected "roster stability" strikes me as an intangible benefit currently weighing in the Blue Jays' favor.

For pitching, I took a similar approach. In 2012 AL teams averaged 1449 innings pitched. For the difference between 1449 and the sum of the innings pitched for each team's projected players, I prorated the stats of all 2012 AL pitchers who pitched 40 or less innings - the types of players likely to serve as fill-ins. Such pitchers gave up 4.94 runs-per-9-innings. I then summed the total projected runs allowed for the players and for the prorated fill-ins, as shown in the following table:

Category	IP	R	R/9
Blue Jays
Projected Players	1425	678	4.28
Fill-ins	24	13	4.88
Total	1449	691	4.29
Yankees
Projected Players	1271	580	4.11
Fill-ins	178	98	4.96
Total	1449	678	4.21

These methods of arriving at full-season team stats could no doubt be tweaked. However, they will do for back-of-the-envelope team projection purposes, especially since the same methods are being equally applied to both teams.

Note: These projections are calculated using a program I wrote to implement the Marcels algorithm (documented here).