Progressing Toward Better Stats Thread

[brett]

Sultan, I like your table. I would only like to see road s-OPS+ added in. If people see that the road s-OPS+ is higher than overall OPS+, or lower, they can see that they got a particular benefit from their park, more than just the park factor, and people can decide for themselves how to factor it in, but you don't have to do anything that people might think is arbitrary.

Your road s-OPS+ work by itself is worth looking at a table for.

[layson27]

...why was Gehrig hurt by YS? I'm interested. Did he have a lower BA, fewer home runs?

From 1925-1939 Gehrig hit 10 more HR's at home than on the road in 323 less PA's, so that can't be it. But his BA was (oddly?) 22 points less and he hit 119 less doubles. More evidence that old YS hurt double totals. That explains Gehrig's H/A slugging difference.

[Sultan_1895-1948]

From 1925-1939 Gehrig hit 10 more HR's at home than on the road in 323 less PA's, so that can't be it. But his BA was (oddly?) 22 points less and he hit 119 less doubles. More evidence that old YS hurt double totals. That explains Gehrig's H/A slugging difference.

Hitters back then didnt hit with as much tilt, like a Dunn or an Edmonds. They couldnt afford to with the zone being high and low.

If you look at Gehigs swing, he bends at the knees during load, really getting his lower body involved. His swing is rather level which allowed him to stay on top of high pitches. He was more of a line drive spray hitter who could occasionally blast one with backspin, but generally his balls didnt tend to be moonshots.

Not sure what Lou's career numbers are at Fenway. While its true it favored righties for obvios reasons, it helped lefties due to the wall increasing oppo margin of error.

Will see what Jenkinson wrote about Lou in his "Ultimate Power" book when I get home.

[Sultan_1895-1948]

Those are right, but you have to adjust for league offense. Anyway, the fact that Gehrig was hurt by YS and slugged .644 on the road is a pretty big deal. It would affect my ranking of him. Williams does as expected. His .615 road slugging in that league environment is right in line with his relative rates. His relative rates were not helped by his park (which favored right handed hitters on the whole anyway), but why was Gehrig hurt by YS? I'm interested. Did he have a lower BA, fewer home runs?

So even though we already included leagueSAeff, since the final step deals with raw SA, league SA needs to be included.

Not much difference between the players in question. Williams league was .409 to Lou's .413.

So we have four numbers to consider.
1. Players raw road SA
2. Players raw difference between their home/road SA
3. Players league SA during their career

4. Their SAeff+

[Sultan_1895-1948]

Ok, still can't log in at home.

Had to email this to myself and download from cell to post on here.

See what you think of this Brett. I called it total slugging value but maybe production would be more fitting.

Basically taking road slugging and applying the expected boost.

Then comparing that number to how their league slugged.

Then factoring in park factors they played with.

Finally, including their SAeff...just theirs vs their leagues. No more park factors since it was already acounted for.

When you add the two numbers together you get something like 2.56. To put into BA form, I divide that by 10, or obviously just slide the decimal.

Sorted by TSV.

Seems to do a decent job of giving or taking away proper credit. Look at Aaron and Foxx for example.

I don't know what "avg" would be though.

By the way, Klein and probably some others would be either hurt, or helped, if there was a min PA required for "that year" to count in park factors. For instance, through '37 his PF was 107 (ends up being 102) and he didnt play a whole lot after that. So his earlier production is made to appear less influenced than it really was.

TSVchart.jpg

[Sultan_1895-1948]

Those are right, but you have to adjust for league offense. Anyway, the fact that Gehrig was hurt by YS and slugged .644 on the road is a pretty big deal. It would affect my ranking of him.

How bout this.

If you remove only his production in Griffith from his road numbers....that leaves six main road stadiums with two (Braves Field and Cleveland Stadium) where he got a combined 138 career at bats. Not to mention, in the 85 AB in Cleveland Stadium, he posted by far his worst slash (.494 SA).

Anyway, just removing Griffith from his road park pool, Gehrig hits .356 and slugs .663 on the road. That .663 is 11 points higher than Williams' inflated HOME SA.

[brett]

I am looking at it. So I may have confused you. The first 4 columns of data are correct.

From there, all you need to do is average the road SA and the expected home SA to get "total expected SA" OR a simple way is to just take road SA and multiply by 1.02 to get expected total SA.

At that point, you actually have to RAISE players with high offensive parks, because their league SA that they are rated against goes up because of their park. However, park factor does not let you convert one slugging percentage to another.

IF the league slugged .400, and the park factor was 1.07, that does not mean that the league would have a .428 slugging percentage. It means that the league would have 1.07 times as much scoring with that park half time. So there is no way to use park factors to adjust slugging percentages.

What you should do is make a data base of the ACTUAL league slugging percentage for each year, (not the league SA that is compiled for the player under the more stats tab). That's all.

So really just 2 steps to get there: take road SA times 1.02 to get expected total SA. Divide by the ACTUAL league SA for the years spanning his career. Technically it should be weighted by the players PAs, but you have to decide if you want to do that.

SINCE slugging efficiency DOES appear to line up with total offense, you CAN divide it by the park factor to adjust out his park. Then divide by league SAEff.

[Sultan_1895-1948]

This would be much easier if I could log in. How are you able to? Have to reach status perhaps lol.

Anyway, so basically instead of multiplying road SA by .04 and adding that to the original, I need to multiply by 1.02.

Not sure we even need to inclide PF now, given what you said.

I have excel files for each leagues OPS+ and SAeff.

I can make one for SA but guess I didn't figure it would be different. Why is it different?

[Sultan_1895-1948]

Here's a chart I made using leage SAeff data.

All of the expected years behave as expected. The expansion years, the cushioned cork center in 1926, 1930, 1968, 1987 etc

AL-NL-SAeffCHART.jpg

[layson27]

So really just 2 steps to get there: take road SA times 1.02 to get expected total SA. Divide by the ACTUAL league SA for the years spanning his career. Technically it should be weighted by the players PAs, but you have to decide if you want to do that.

I just want to point out that the league slugging averages from post#106 are the actual league numbers. Example: League SA during Ruth's career was .404 & during Williams' career the Al average was .398 whereas the park adjusted lg avg. was .409.

[Sultan_1895-1948]

Woohoo!!!

The lockout is over!!!

[Sultan_1895-1948]

I just want to point out that the league slugging averages from post#106 are the actual league numbers. Example: League SA during Ruth's career was .404 & during Williams' career the Al average was .398 whereas the park adjusted lg avg. was .409.

Not using the leagueSA number shown in the "More Stats" section, I have Ruth's leagues at .381 avg.

I still don't know why there would be such a big difference in some years. Never got an answer.

[layson27]

Not using the leagueSA number shown in the "More Stats" section, I have Ruth's leagues at .381 avg.

BBRef lists Ruth's park adjusted league slugging average (lgSLG) for his career at .401. I think .404 is right for the raw SA.

[Sultan_1895-1948]

I used this http://www.baseball-almanac.com/hitting/hislug3.shtml to enter into excel.

Brett said not to use the LGslg listed under the "more stats" at bbref.

Just as an example, 1920 shows .407 for that one year under Ruth's "more stats" page. But if you actually click on AL, it shows the league slugged .387.

[Sultan_1895-1948]

On a side, note I was just looking through some of these guys' actual SA versus their "expected" overall SA. If the difference from expected to actual goes down, they are in red. And obviously if their actual was higher it's in blue.

Aside from the platoon boost (which I think righties should get something at least), there has to be a way to account for this, beyond just simple park factors.

Sorted by EXPECTED OVERALL, look at Piazza and Foxx. They're both gonna have leagues with a good SA; they are about dead even, except heading in the opposite direction; like a 75 point gap!

Or check Mathews sandwiched between Greenberg and Mize. Then look over to how that's impacting his SAeff.

DIFFbetweenEXPandACTUAL.jpg