1996-2005. You don't want to look, like Scrooge when confronted with the ghost of Christmas future. 

But we have to. Ugh. 160. 


In 1999 alone 22 players had 500 PA, .400 and higher OBP and SLG of at least .500. 22,  MORE than in all of  1976-1985 put together. 

In 2000 there were 18. 

Bonds' .863 is of course the leader of the pack for SLG. And he hits over .800 SLG again in 2004.  Seven other players top .700.  Another 53 top .600. That is as many as topped .500 in 1956-1965 and 1976-1985 altogether, 20 seasons. Sigh. 

Edgar Martinez and Gary Sheffield each had seasons where their OBP was within 98 points of their SLG.  Brian Giles had a year where he was withing 87 points. That is the closest ratio I see between an OBP and SLG in this era. Pretty much OBP is a form of fear factor - pitchers would rather walk you than pitch to you. Period. Not much strategy involved. Try to keep the ball in the park.

...and in 2006-2015, just like stocks, they come back down - 72, which seems normal after what we just went through for ten years. Only 7 players a year with 500 PA; .400 or higher OBP and .500 plus SLG: 

13 are in 2006, so we still have a hangover from the steroids. 17 in 2007.  Back down to 9 in 2008 and in 2009 again the same. 

And then...in 2010 four, just 4 - Miguel Cabrera, Joey Votto, Albert Pujols and Josh Hamilton. 

2011 - 6
2012 - 2 - Buster Posey and Prince Fielder
2013 - 4
2014 - 3
2015 - 5 - Goldschmidt, Harper, Votto, Trout, Cabrera

So, I think that the findings are consistent with the idea that OBP goes up and down with SLG/HRs, and that SLG is arguably the most important stat. I tried to control for this already by checking out the frequency of .400 OBP but SLG below .400 and then between .400 and .500.

I think we found that: 

OBP had a certain independence in the 19th and early 20th century - half the great seasons were driven by batting average that led to most of the high OBPs, but half were OBPs well higher than the batting average, without SLG being very high. - 23 of 46 for the 1885-1915 period. 

After that we find that high OBP seasons are either mainly a factor of high SLG and home runs, or are the product of 1-2 great players having great seasons in which they really do manage to get on base well more than than their batting averages or slugging percentages (and subsequent pitcher fear of them hitting a long ball) could explain. Some of these, like Joe Morgan and Tim Raines are even counter-intuitive - guys that were such good base stealers that the risk of them hitting a double or homer was less than them stretching a walk to second or even third base, but who still got on base more than pretty much anyone else. Rickey, though his SLG usually is in the mix might be in that elite camp too. 

Others, like Max Bishop or Eddie Stanky, got on base again without very high averages or home run power and were not base-stealing threats. 

But these players have not only been few and far between but as we have seen nearly non-existent except for these exceptional players in their best seasons since 1961. That's a long time. 

BUT, if we find that there are a significant number of players with OBP under .400 but with SLG over .400 and over .500 we might have to admit that  high SLG is a necessary but not a sufficient condition for high OBP, leading us to need to ask some other questions about what else could be factoring in. 

Let's see the numbers: 

Here first is the decade by decade breakdown for players with 500 PA, BELOW .400 OBP, and .500 or ABOVE SLG. 

1885-1895 - 13 - all but one a .300 hitter. 
1896-1905  - 13 again - all .300 hitters
1906-1915 - 7  none before 1911, and only one not a .300 hitter - and he's Gavvy Cravath
1916-1925 - 30  five below .300 AVG but they all hit a lot of HRs
1926-1935 - 70 - I admit at this time that I am perplexed - take Al Simmons 1929 - how do you get a .642 SLG, hit .365 with 34 HRs and only have a .398 OBP?  But appearances may be deceiving: the lowest OBP of this group is Joe Medwick at .343 with .529 SLG.  Maybe his OBP IS still driven by SLG? But he hit .319.  I have to do the math later

1936-45 - 46
1946-55  82 !!
1956-65 -  READY?  126  
1966-75 - 90
1976-85 - 137
1986-1995 - 131
1996-2005 - 331  !!!
2006-2015 - 242

Okay, I am going to have to look closely at some cases and crunch the numbers to understand this better. Not all of it can be attributed to high strikeout rates, for example, since those are low in the 1930s and 40s and 50s even among HR hitters compared with today. But the more recent stuff may have a high level of strikeouts involved in such high SLG with such low OBP. But that can't explain all of it. 

We seem to find that SLG is a necessary but not a sufficient condition for a high OBP, but that means that other factors are involved and since we see enormous differences between decades, we may be able to identify some of the variables involved. but I need to think about these results and need to get some - ahem, work done for an hour or so, followed by Italy dinner time. So stay tuned. Same batting time, same batting station. 
 

Posted by dahsdebater on 1/10/2016 6:01:00 PM (view original):
Italyprof, based on your analysis, and assuming that correlation implies causation (which is completely ridiculous, as you well know), how are we to infer that home runs promote better OBP, rather than better OBP resulting in greater HR totals?

Higher OBP results in more plate appearances, which results in more chances to hit home runs.  If all you're looking at is raw HR totals, that matters.  Pitchers also tend to be less effective, on average, with runners on base.

dahsdebater, I apologize for not replying to this sooner - I didn't see it until now when I logged on and to refresh my middle aged brain about what the heck I had been writing about I re-read this thread from the start to find your note. I think this got lost in one of the moments when we posted at the same moment or else my compute froze. 

First, thank you for the question - I am not claiming to have answers but to be testing hypotheses that are falsifiable, some are mine and some are sacred cows as it were that have gone unchallenged now for a few years. 

Your question is a good one and you might be right. 

But I think that my having worked my way through the evidence to find SLG more important and not just raw HR numbers helps strengthen this analysis. I am not sure that the effect you are referring to is really likely to be that great even if it were to be shown empirically by the data, but you at least have a plausible counter-argument to present here. But slugging is not a raw number and while we could argue that higher OBP by definition leads to more total bases and therefore higher SLG, I think the march through the decades here has shown that while one can have a high slugging percentage without having a high OBP, you can't really, since maybe Max Bishop's time, and as a common occurrence since the deadball era, have a high OBP without a high SLG, unless you are an elite player having a great year (Joe Morgan in 1975, etc.). 

Also, here is a table from baseball-reference (I have learned now how to use the site to custom-make statistical tables with their data - in which all baseball seasons are listed from top to bottom in order of  how high OBP was: 

Year	R/G	HR	BB	SO	BA	OBP ?	SLG
1894	7.44	0.40	3.70	2.10	.309	.379	.435
1895	6.60	0.31	3.22	2.27	.296	.361	.400
1930	5.55	0.63	3.10	3.21	.296	.356	.434
1893	6.57	0.29	3.91	2.13	.280	.356	.379
1925	5.13	0.48	3.15	2.71	.292	.354	.411
1897	5.89	0.23	2.92	2.30	.292	.354	.386
1896	6.04	0.26	3.06	2.22	.290	.354	.387
1929	5.19	0.55	3.25	2.84	.289	.353	.417
1936	5.19	0.55	3.40	3.33	.284	.349	.404
1924	4.76	0.36	2.99	2.70	.287	.348	.394
1922	4.87	0.43	2.93	2.81	.288	.348	.401
1921	4.85	0.38	2.79	2.83	.291	.348	.403
1923	4.81	0.40	3.07	2.85	.284	.347	.391
1950	4.85	0.84	4.02	3.86	.266	.346	.402
2000	5.14	1.17	3.75	6.45	.270	.345	.437
1999	5.08	1.14	3.68	6.41	.271	.345	.434
1927	4.75	0.37	3.01	2.79	.284	.345	.393
1926	4.64	0.35	3.12	2.75	.281	.345	.389
1949	4.61	0.69	4.04	3.61	.263	.344	.384
1939	4.82	0.59	3.44	3.45	.275	.344	.397
1928	4.73	0.44	3.11	2.88	.281	.344	.397
1938	4.89	0.60	3.53	3.41	.274	.343	.396
1937	4.87	0.58	3.41	3.63	.277	.343	.399
1899	5.25	0.19	2.70	2.09	.282	.343	.366
1934	4.91	0.55	3.21	3.45	.279	.342	.397
1948	4.58	0.63	3.89	3.64	.263	.341	.382
1935	4.90	0.54	3.19	3.26	.279	.341	.397
1996	5.04	1.09	3.55	6.46	.270	.340	.427
1994	4.92	1.03	3.48	6.18	.270	.339	.424
1931	4.81	0.43	3.11	3.19	.278	.339	.391
1900	5.21	0.22	2.67	2.36	.279	.339	.366
1995	4.85	1.01	3.53	6.30	.267	.338	.417
2006	4.86	1.11	3.26	6.52	.269	.337	.432
1997	4.77	1.02	3.46	6.61	.267	.337	.419
1932	4.91	0.55	3.06	3.19	.277	.337	.400
1912	4.53	0.18	3.12	3.93	.269	.337	.359
1890	6.01	0.24	3.64	2.04	.260	.337	.351
2007	4.80	1.02	3.31	6.62	.268	.336	.423
1953	4.61	0.84	3.50	4.12	.264	.336	.397
1951	4.55	0.75	3.73	3.77	.261	.336	.386
1947	4.36	0.63	3.71	3.68	.261	.336	.377
1911	4.51	0.21	3.17	3.99	.266	.336	.357
2004	4.81	1.12	3.34	6.55	.266	.335	.428
1998	4.79	1.04	3.38	6.56	.266	.335	.420
1920	4.36	0.26	2.76	2.94	.276	.335	.372
1941	4.49	0.53	3.57	3.55	.262	.334	.375
1940	4.68	0.64	3.35	3.66	.267	.334	.392
1898	4.96	0.16	2.76	2.29	.271	.334	.347
1889	5.96	0.31	3.36	3.52	.264	.334	.358
2009	4.61	1.04	3.42	6.91	.262	.333	.418
2008	4.65	1.00	3.36	6.77	.264	.333	.416
2003	4.73	1.07	3.27	6.34	.264	.333	.422
1954	4.38	0.78	3.65	4.13	.261	.333	.390
2001	4.78	1.12	3.25	6.67	.264	.332	.427
1993	4.60	0.89	3.33	5.80	.265	.332	.403
1955	4.48	0.90	3.66	4.38	.259	.332	.394
1887	6.34	0.29	2.86	2.80	.271	.332	.374
2002	4.62	1.04	3.35	6.47	.261	.331	.417
1987	4.72	1.06	3.42	5.96	.263	.331	.415
1956	4.45	0.93	3.63	4.64	.258	.331	.397
1891	5.69	0.26	3.62	3.45	.254	.331	.343
2005	4.59	1.03	3.13	6.30	.264	.330	.419
1979	4.46	0.82	3.24	4.77	.265	.330	.397
1933	4.48	0.44	3.00	3.04	.270	.330	.376
1977	4.47	0.87	3.27	5.16	.264	.329	.401
1945	4.18	0.41	3.37	3.27	.260	.329	.355
1961	4.53	0.95	3.46	5.23	.258	.328	.399
1946	4.01	0.49	3.54	3.90	.256	.328	.360
1975	4.21	0.70	3.46	4.98	.258	.327	.374
1952	4.18	0.69	3.54	4.19	.253	.327	.370
1901	4.99	0.20	2.46	3.14	.272	.327	.360
1986	4.41	0.91	3.38	5.87	.258	.326	.395
1980	4.29	0.73	3.13	4.80	.265	.326	.388
1970	4.34	0.88	3.53	5.75	.254	.326	.385
1962	4.46	0.93	3.37	5.42	.258	.326	.393
1944	4.17	0.42	3.19	3.30	.260	.326	.358
2010	4.38	0.95	3.25	7.06	.257	.325	.403
1990	4.26	0.79	3.29	5.67	.258	.325	.385
1983	4.31	0.78	3.20	5.15	.261	.325	.389
1973	4.21	0.80	3.37	5.24	.257	.325	.379
1958	4.28	0.91	3.29	4.95	.258	.325	.394
1913	4.04	0.19	2.94	3.76	.259	.325	.345
1982	4.30	0.80	3.16	5.04	.261	.324	.389
1974	4.12	0.68	3.33	5.01	.257	.324	.369
1960	4.31	0.86	3.39	5.18	.255	.324	.388
1959	4.38	0.91	3.31	5.09	.257	.324	.392
1957	4.31	0.89	3.31	4.84	.258	.324	.391
1991	4.31	0.80	3.32	5.80	.256	.323	.385
1985	4.33	0.86	3.29	5.34	.257	.323	.391
1984	4.26	0.77	3.16	5.34	.260	.323	.385
1978	4.10	0.70	3.23	4.77	.258	.323	.379
1943	3.91	0.37	3.38	3.45	.253	.323	.344
1942	4.08	0.44	3.43	3.40	.253	.323	.350
1992	4.12	0.72	3.25	5.59	.256	.322	.377
1919	3.88	0.20	2.68	3.06	.263	.322	.348
1902	4.44	0.16	2.44	2.97	.267	.322	.344
2011	4.28	0.94	3.09	7.10	.255	.321	.399
1914	3.86	0.19	2.96	3.93	.254	.321	.337
1989	4.13	0.73	3.21	5.61	.254	.320	.375
1981	4.00	0.64	3.18	4.75	.256	.320	.369
1976	3.99	0.58	3.20	4.83	.255	.320	.361
1969	4.07	0.80	3.45	5.77	.248	.320	.369
2012	4.32	1.02	3.03	7.50	.255	.319	.405
2013	4.17	0.96	3.01	7.55	.253	.318	.396
1988	4.14	0.76	3.09	5.56	.254	.318	.378
1915	3.81	0.17	2.98	3.76	.250	.318	.332
1910	3.83	0.14	2.97	3.87	.249	.318	.326
2015	4.25	1.01	2.90	7.71	.254	.317	.405
1971	3.89	0.74	3.23	5.41	.249	.317	.365
1918	3.63	0.12	2.83	2.92	.254	.317	.325
1903	4.44	0.15	2.41	3.55	.262	.317	.346
1892	5.10	0.23	3.35	3.24	.245	.317	.328
2014	4.07	0.86	2.88	7.70	.251	.314	.386
1964	4.04	0.85	2.96	5.91	.250	.313	.378
1916	3.56	0.15	2.84	3.82	.248	.312	.326
1871	10.47	0.19	1.55	0.69	.287	.312	.384
1972	3.69	0.68	3.15	5.57	.244	.311	.354
1965	3.99	0.83	3.09	5.94	.246	.311	.372
1917	3.59	0.13	2.77	3.48	.249	.311	.324
1966	3.99	0.85	2.89	5.82	.249	.310	.376
1963	3.95	0.84	2.96	5.80	.246	.309	.372
1905	3.89	0.14	2.50	3.85	.248	.307	.323
1967	3.77	0.71	2.98	5.99	.242	.306	.357
1909	3.54	0.10	2.62	3.78	.244	.306	.311
1906	3.61	0.11	2.51	3.71	.247	.306	.314
1907	3.52	0.10	2.47	3.58	.245	.305	.309
1873	8.99	0.12	0.84	0.70	.290	.304	.357
1886	5.47	0.20	2.64	2.02	.246	.303	.332
1904	3.72	0.13	2.23	3.72	.247	.301	.321
1968	3.42	0.61	2.82	5.89	.237	.299	.340
1908	3.38	0.11	2.36	3.65	.239	.297	.305
1872	9.26	0.10	0.72	0.72	.285	.297	.348
1888	4.87	0.24	2.16	3.76	.239	.291	.320
1881	5.10	0.11	1.54	2.65	.260	.290	.338
1877	5.67	0.07	0.96	2.02	.271	.289	.338
1885	5.22	0.18	1.98	1.87	.244	.288	.325
1883	5.75	0.15	1.48	1.83	.257	.286	.345
1874	7.48	0.09	0.51	0.77	.273	.282	.333
1884	5.42	0.22	1.60	1.40	.243	.279	.327
1878	5.17	0.06	0.99	2.94	.259	.279	.319
1876	5.90	0.08	0.65	1.13	.265	.277	.321
1882	5.33	0.16	1.39	1.89	.248	.276	.330
1879	5.31	0.09	0.79	2.87	.255	.271	.329
1880	4.69	0.09	1.09	2.93	.245	.267	.320
1875	6.14	0.06	0.36	0.98	.254	.261	.310

Look up 1995 on that list above: it is right after 1900 at .341 OBP. It is way the heck down the list. Before 1995 on this list,  the only steroid and post-steroid era seasons on this list - again ranked in order of season major league-wide OBP - are 2000 and right after it 1999. The season 1950 ranks higher. That is not counting 1994 which is a shortened season and in any case does not rank much higher either.

So how can higher OBP lead to higher HRs if most of the high OBP seasons are not the ones with the most home runs, AND more to the point, how can we be living in an era of higher OBP if OBP is NOT higher than it used to be? 

To be clear, I don't really believe that OBP is a major causative factor for HRs or SLG.  It probably does impact the raw HR totals a little bit, for the reason cited in my original post, but that's still something you can avoid by using HR rates, which are widely available.  I ALSO don't think SLG is a major causative factor for OBP, though it helps a bit with walk totals.  I don't think there is a particularly strong causative relationship here.  I think there is a correlation/causation confusion here.  Both OBP and SLG are undeniably related to, in essence, being a good hitter.  Generally speaking, in eras when hitters have the advantage in matchups relative to the historical average, both tend to go up.  When pitchers are dominating - as now, and in the '60s - both go down.  For a league-average hitter, batting average is a fairly dominant chunk of both OBP and SLG, so when looking at league data you have to figure there's an inevitably substantial correlation between OBP and SLG through AVG.  It might be more informative to at least study the relationship between (OBP-AVG) and Iso.  But again, I would expect it to still be related based on hitter quality and relative advantage.  In particular, getting deeper into counts tends to result in a greater rate of XBH, and obviously all plate appearances resulting in walks are long.  I don't think the average plate appearance has ever been 4+ pitches, so any walk is a long PA.  But none of this implies to me that OBP causes SLG or that SLG causes OBP.

With regards to this specific question "So how can higher OBP lead to higher HRs if most of the high OBP seasons are not the ones with the most home runs?" I think you might be confusing some factors too much.  Most of the highest OBP seasons are deadball seasons.  It should be inherently obvious why those are not high HR seasons.  Furthermore, for the first few decades of the modern era, HRs were still not "in vogue" as they are today.  It was in the '60s when it started to become more popular to be bigger and stronger to hit HRs, but it seems pretty obvious that the reason baseball had its steroid scandal literally decades after football was because the theory that you lost too much bat speed and reaction time by being too muscular persisted into the '90s.  In addition to the "cultural" factors, I would also refer back to the fact that I would only expect OBP to be a minor causation factor in HR totals.  Rates are probably more important than the relatively small variance in number of opportunities during the modern era.

During the modern (live-ball) era, OBP only spans the range from .299 to .354.  If you assume that all games are 27 outs (that is, no skipped bottom of the 9th, and no double plays), then an average game in 1930 contained ~41.8 plate appearances per team.  A game in 1968 consisted of 38.5 PAs per team.  The difference of 3.3 is less than 10%.  It's not nothing, but compared to the variance in HR rates throughout history it probably doesn't matter.

The long and the short of it is, if you look at individual player-seasons with 500+ PAs during the modern era, the R^2 correlation between OBP and SLG is .434.  They're related, but maybe not enough to be considered likely major causative factors in one another.  The correlation of OBP to average is of course substantially stronger (.560), as hits are a major driver of times on base.