Carew vs Morgan

[Jackaroo Dave]

The OB/SLG/Ops+ for Joe Morgan and Rod Carew are uncannily similar, .392/.427/132 and .393/.429/131 respectively. Yet they reached these numbers by almost opposite routes: Carew hit lots of singles and Morgan did lots of everything else. It's hard to think of two players who are so similar in some senses yet so different in others.

Anyway, this combination dovetails exactly with the issues we've been discussing in a couple of threads on evaluating different types of hitters with various tools. So I thought it might be revealing to compare these two in seasonal notation.

.....PA...AB...H....1B..2B.3B.HR.SB.CS..BB..BA..OB A..SLG.OPS+.TB.ISOP.WAR.ADJWAR
M. 693 561 154 105 27 6 16 42 10 114 .271 .392 .427 132 242 .156 6.67 6.28*
C. 692 611 200 158 29 7 6 23 12 67 .328 .393 .429 131 262 .101 5.11 5.30*

Sorry for the crappy layout. Even so, the similarities and differences jump off the page. Also, These are transcribed, so there may well be minsprits. Best to get the stats from BBreference directly.

*Offensive WAR includes steals and caught stealing, so I peeled them off using linear weights of .2 for SB and -.435 for CS. They are rough and there are other figures around, but I don't think they make too much difference. The offensive WAR is just in there for quick and dirty reference anyway.

So, what do you think? Morgan is the sabremetric poster child, and Carew got his runs the unfashionable way, but when they were playing, it was a different story.

The one item that cauught my eye was the identical OPS+ score, considering that their OB and SLG were identical, but Morgan played through the sixties, and in the Astrodome at that, while Carew did not. Something funny going on with OPS+?

[willshad]

This is actually an interesting thread. Morgan is considered a top 15 or so player by many in the sabermetric community, while Carew is merely a 'solid hall of famer'. While I am sure that most of their 'advanced' stats show Morgan to be significantly more valuable for his career, I just don't see it. Like you have shown, their hitting is almost identical. I give Carew the advantage, though, because his OPS+ is much less walk-heavy. Morgan does have an advantage on the bases (though Carew was no slouch there either). Carew also has an advantage in the field. I think Carew also sustained his prime longer..7 batting titles, and a couple of near misses. His 1977 holds up to any year Morgan put up.

I think the 'Astrodome disadvantage' is way overblown , especially when it comes to Morgan and Jim Wynn. They both hit better at the Astrodome than on the road in most seasons...thus, it actually worked in their advantage to play there, because most guys hit poorly there. I don;t know for sure, but I suspect it was the power guys who were hurt by that park, and speed guys were helped by it.

In any case, I think Morgan and Carew and very close to being equals for career value.

[Jackaroo Dave]

Like you have shown, their hitting is almost identical. I give Carew the advantage, though, because his OPS+ is much less walk-heavy. .

What's interesting to me is that there are two schools of thought here, yours, which gives Carew the nod BECAUSE his OPS+ is less walk-heavy, and the other which gives Morgan the nod BECAUSE his OPS+ is more walk heavy. So far, I have . . .
Favoring Carew:
Basic runs created (TB * OBA) 103-95, because their on base average is the same, but Carew's singles yield 20 more TB.
Leewileyfan's runs created estimator (BB + TB) * BA, 108-97, despite Morgan's big edge in walks, because of the 56 point BA difference.

Favoring Morgan: Baseball Reference's version of Runs Created, 110-105. See the formula there for details, but it no doubt includes sb, cs, sf, and gidp, where Morgan has the edge.
Baseball Reference's version of offensive WAR, 6.67-5.11 including sb,cs; 6.26-5.65 with only plate offense.

Washes:

Obviously, SLOB (SLG *OBA) and OPS, since their rates are identical.
Less obviously, OPS+ and other BB-Ref relativized offensive measures like neutralized OPS.

Still out there . . . wOBA, linear weights, base runs, other higher-hanging fruit.

[tag0519]

Nice thread. Thanks Jackaroo and nice research. Oh by the way, don't mean to start anything here, but what's wrong with a walk? It is a skill. Again, interesting topic on two of my favorite players growing up.

[BigRon]

What hasn't been mentioned is Morgan's very significant edge in power. Morgan's isolated power (ISO) is .156, well above league averages, while Carew's is .101, well below average. They had similar 2B and 3B rates, but Morgan is way ahead in homers. Basically, Morgan was a threat to hit the ball out of the park, Carew wasn't.

[willshad]

What hasn't been mentioned is Morgan's very significant edge in power. Morgan's isolated power (ISO) is .156, well above league averages, while Carew's is .101, well below average. They had similar 2B and 3B rates, but Morgan is way ahead in homers. Basically, Morgan was a threat to hit the ball out of the park, Carew wasn't.

Their slugging percentages, as well as relative slugging percentages, were almost identical. Whatever home run/walk advantage Morgan had was more than compensated for by Carew's singles and triples advantages. Morgan makes up for this with his steals. Overall, offensively, they are practically the equal, with maybe a SLIGHT edge to Morgan.

[Bothrops Atrox]

The 20 WAR gap can almost all be explained by positional differences. Carew played a majority of his games at 1B and Morgan played almost his whole career at 2B. Morgan is ahead offensively (mostly because of a 5 WAR gap due to baserunning) and Carew is ahead in terms of run prevention. They have very similar "replacement" value. If buy into the positional scarcity philosophy, their WAR gaps make sense.

If you don't buy into the positional difference issue, than I wouldn't even bother considering WAR at all.

[brett]

If we are talking just about hitting, Morgan had 491 batting runs in 2649 games, and Carew 430 in 2469 games (versus average).

That would be .185 for Morgan and .174 for Carew per game.

I will point out that Morgan (as a lot of lower BA, walk heavy on-base guys are) was .226/.390/.345 with a runner on second or second and third with first base empty and Carew is approximately .335/.460/.450. These are situations where the single to walk value ratio goes much higher than 1 to 0.7.

[Ubiquitous]

Joe Morgan had 1216 PA with a runner on second or a runner on second and third. He had 227 RBI in those scenarios while committing 734 outs.

That is 18.7 RBI and 60.4 outs per 100 PA.

Rod Carew had 1151 PA in the same situations. He had 272 RBI and 615 outs.

That is 23.6 RBI and 53.4 outs per 100 PA.

Carew had 122 IBB to Morgan's 72 IBB. Over 10% of Carew's PA with first base open were of the intentional walk variety. So a good chunk of Carew's game in these scenarios was in fact the walk.

Now then Morgan had 824 RBI with RISP and Carew had 859.

Morgan is at 30.6 RBI per 100 PA and Carew is at 32.8 RBI per 100 PA. So the nearly 5 run lead is cut by over half when looking at all RISP scenarios.

Then Joe is at 2.4 RBI per 100 PA with bases empty to Carew's .8 RBI per 100 PA

Joe is at 7.1 RBI per 100 PA with a runner just on first. Carew is at 5.5 RBI per 100 PA

So while in a specific situation that happens less often than almost all other scenarios Carew is better but in situations that happen most often Morgan is better.

[BigRon]

Their slugging percentages, as well as relative slugging percentages, were almost identical. Whatever home run/walk advantage Morgan had was more than compensated for by Carew's singles and triples advantages. Morgan makes up for this with his steals. Overall, offensively, they are practically the equal, with maybe a SLIGHT edge to Morgan.

Yeah, but slgging average isn't the best indicator of power- isolated power (ISO) is far more informative. Slugging average takes into account ALL bases from ALL hits, so a high average singles hitter like Carew comes out with a respectable SA even though he doesn't have a lot of power. ISO only uses extra bases on hits.

[willshad]

Yeah, but slgging average isn't the best indicator of power- isolated power (ISO) is far more informative. Slugging average takes into account ALL bases from ALL hits, so a high average singles hitter like Carew comes out with a respectable SA even though he doesn't have a lot of power. ISO only uses extra bases on hits.

I know this..Morgan does indeed have more home run power. My point is that Carew compensates for that with his huge advantage in batting average.

[brett]

Joe Morgan had 1216 PA with a runner on second or a runner on second and third. He had 227 RBI in those scenarios while committing 734 outs.

That is 18.7 RBI and 60.4 outs per 100 PA.

Rod Carew had 1151 PA in the same situations. He had 272 RBI and 615 outs.

That is 23.6 RBI and 53.4 outs per 100 PA.

Carew had 122 IBB to Morgan's 72 IBB. Over 10% of Carew's PA with first base open were of the intentional walk variety. So a good chunk of Carew's game in these scenarios was in fact the walk.

Now then Morgan had 824 RBI with RISP and Carew had 859.

Morgan is at 30.6 RBI per 100 PA and Carew is at 32.8 RBI per 100 PA. So the nearly 5 run lead is cut by over half when looking at all RISP scenarios.

Then Joe is at 2.4 RBI per 100 PA with bases empty to Carew's .8 RBI per 100 PA

Joe is at 7.1 RBI per 100 PA with a runner just on first. Carew is at 5.5 RBI per 100 PA

So while in a specific situation that happens less often than almost all other scenarios Carew is better but in situations that happen most often Morgan is better.

You demonstrate that Morgan is better, but the splits still make them closer to each other than if their productivities were random.

[dominik]

The 20 WAR gap can almost all be explained by positional differences. Carew played a majority of his games at 1B and Morgan played almost his whole career at 2B. Morgan is ahead offensively (mostly because of a 5 WAR gap due to baserunning) and Carew is ahead in terms of run prevention. They have very similar "replacement" value. If buy into the positional scarcity philosophy, their WAR gaps make sense.

If you don't buy into the positional difference issue, than I wouldn't even bother considering WAR at all.

this and morgan had about 800 more PAs and 200 more games. so basically 1.5 more seasons to collect WAR.

[Ubiquitous]

You demonstrate that Morgan is better, but the splits still make them closer to each other than if their productivities were random.

No, one split indicates that Carew was better in one area while 7 other splits indicate that Morgan was better than Carew in those splits.

[leewileyfan]

It is an interesting match up. What I offer below [FWIW] is my own approach to comparing players as far as overall career production is concerned:

1. I settle on a particular batting runs created metric, in this case Baseball-Reference Player Page; Advanced Batting; Runs Created.

2. I relate offensive production as a four decimal place number, the quotient produced by dividing those RC by Plate Appearances. For batting "seasons," I consider 550 PA a "season."

3. I select, if it's an overall evaluation, THE primary position played IF that is commonly shared by players being compared. In the case of Morgan and Carew I use 2B.

4. I separately, using my own defense metric, determine a defense rating at the position for each player. I use innings played to determine playing time games. I consider 150 games to be a "season" at the position.

5. If a player has an average of +/- 5 Defense Runs [compared to "average" per season, and the playing time = 6.7 "seasons," he gets the product of DR * seasons ... +5 * 6.7 = =33.5; -5 * 6.7 = -33.5]. That's the total of his defense runs ... any other defensive play is ignored, as if he were a DH.

6. I add the DR to the batting RC and divide the total by Plate Appearances. That, to me, indicates net productivity in which position-focused defense runs do get diluted to the dominant factor of batting performance. I am OK with this in that we are attempting to evaluate position players per PA; and batting opportunities and failures are more dominant numerically than defensive opportunities.

Some of the results at first reading [to me] are harsh. Of over 700 players I have on the spreadsheet, over 150 produce NEGATIVE run production [or saving] numbers when compared to average. The standard I use for "average" is .12 RC per PA, such that a batter with 72 RC and +/ "0" defense runs would be exactly average. He would be +0 ... which might constitute a 20 year career, with net "0" runs +/- and "0" win credits.

Many HoF players are in the "0" and minus categories, especially those noted for defensive play.

PLAYER ... RC ....... PA ....... DR [2B] ... Productivity/PA ... Runs 600 PA

Morgan ... 1804 .... 11329 ... 3.14 ....... .1595........... .1595*600= 95.7

Carew .... 1595 .... 10550 ... 14 ........ .1525 ............. .1525*600= 91.5

.................................................. .............Carew, 1B = .1564*600= 93.84


EDITED: Posting to another thread on Joe Morgan defensive play, a final detailed review of Morgan's career defense [against MLB AVERAGE] reduced his DR from +84 to a bit above +3.