Neutralized batting - BBRef

[1905 Giants]

So what do you guys think of this "stat"? Do you think its pretty useful or just a kind of an amusement thing?

[willshad]

I've seen people use it to try and show, for example, that Al Kaline was a better hitter, in context, than Al Simmons. This is a suspicious conclusion at best. Just because you were a better hitter in context, that does not mean you are a better hitter. it's kind of like saying that a guy who finishes second in a high school race is better in context than a guy who finishes fifth in the Olympics. Just because he did better in his own context, that doesn't make him faster, all things being equal. And, no, I'm not trying to say that the 1930s were a 'higher level' of competition than the 1960s. I'm trying to say that the higher the bar is set in your own era, the more difficult it is to be great, comparatively speaking,

[Jackaroo Dave]

But if the question is "Which one is greater in his own era?" does Neutralized Batting provide a reliable answer?

Has anyone taken a look inside? I'd be curious to see how it handled different sorts of variation from the mean. For example, both Sam Crawford and Frank Howard played in eras that depressed batting production. But Howard hit, what, five times as many home runs. How can a single set of formulas neutralize that and make everything else come out right between them?

Not saying it couldn't, just asking.

[brett]

But if the question is "Which one is greater in his own era?" does Neutralized Batting provide a reliable answer?

Has anyone taken a look inside? I'd be curious to see how it handled different sorts of variation from the mean. For example, both Sam Crawford and Frank Howard played in eras that depressed batting production. But Howard hit, what, five times as many home runs. How can a single set of formulas neutralize that and make everything else come out right between them?

Not saying it couldn't, just asking.

I do not believe so. A guy could have been a relatively great home run hitter in the deadball era, but since home runs accounted for so little value in that period it is wrong to think that that would translate to someone outhomering the league in a live ball era. Let's just say that a guy hits 6 home runs a year in a league where the average player produces 2 home runs a year. That's no where near equivalanet to hitting 50 a year when an average guy is hitting 17-18 because the league that averaged 2 home runs per player did not emphasize the home run, and did emphasize other things much more.

If we could take a players relative hitting value there might be a reasonable way to divide that value among different stats though.

[Tyrus4189Cobb]

WHat is it exactly?

[Jackaroo Dave]

From a player page in Baseball-Reference, if you click the blue tab next to the red batting tab to get Advanced Batting stats, then scroll down the new page, you'll find it. It converts the player's career and 162 game numbers into their "equivalents" in an historical average context. You can also choose a context, like the late nineties, the sixties,l etc. Or you can specify a league average run per game of your own choosing.

The actual formulas are somewhere on the site, I think, but not easily found.

[brett]

From a player page in Baseball-Reference, if you click the blue tab next to the red batting tab to get Advanced Batting stats, then scroll down the new page, you'll find it. It converts the player's career and 162 game numbers into their "equivalents" in an historical average context. You can also choose a context, like the late nineties, the sixties,l etc. Or you can specify a league average run per game of your own choosing.

The actual formulas are somewhere on the site, I think, but not easily found.

It does effectively give you a player's stats, in his era if he played in a neutral park, and if the total run scoring was standardized. It does not adjust for changes in the game. For example Cobb hits 126 home runs but it is using deadball rates of home runs.

As long as home runs and such for the league are in roughly the same proportion, it works. Foxx for example would have hit about .303 with 501 home runs if we simply adjust his productivity to a neutral park with a 715 run setting, virtually identical to Frank Thomas at .299 with 527 home runs.

I think Cobb's .375 average is about right, but 126 home runs is at a deadball proportion. Home runs only account for about 2.5% of total value in the deadball era and something like 16% today. Adjusted for league home run rates he would have hit 744 home runs, but of course since home runs were not as important back then, the league did not contain as many "home run hitters". Plotting the best fit quadratic using these 2 values (126 and 744) I get 370 home runs which seems reasonable to me.

[Tyrus4189Cobb]

I would be very wary of anything that attempts to put a dead-ball player's home runs into the context of the modern era. Home runs simply weren't the style of the game. A guy who hits five home runs in 1911 may have had more if he used a power game approach. Converting his 5 home runs to modern context wouldn't be right. I think it's better just to say that 10 home runs in 1909 equals 10 home runs in 1909.

[brett]

I understand the complications, however I still like to think about it. Ruth averaged 14 home runs per 488 at bats prior to 1919, and 45 in 488 per season from 1920-1932. If a guy who was trying to mash the ball all the time basically tripled his rates, I think Cobb might have roughly tripled his totals too which would put Cobb at about 350 without any change in approach. It also puts him around 16-29 a year (per 154 games) from '07-'18. Then again he didn't hit many more with the live ball. If we give him the incedental 3.2 x before 1919 he finishes with about 260

Also Cobb's average was even more amazing in context. He hit .366 in a league with a .280 BABIP. Hornsby's league put up about a .295 BABIP. Give him a BABIP proportionately higher, OR give him those 350 home runs and he bats a little over .385. Obviously some of the BABIP boost was due to hitting the ball harder, but BABIP also dropped some due to more home runs, but if Cobb had just received the same incidental home run boost that Ruth did (assuming the same basic approach, and got ANY boost from BABIP he would have hit .378 relative to Hornsby's .358 in the same setting.

I don't have a clue how they get Hornsby with a neutralized average of .361. His league/park's adjusted BA was .282 which would put him at 1.28 relative BA. Pujols' league has hit .264 and he .325 which is 1.23 relative. and they have him at .325 neutralized. I understand that batting average accounts for far less of players' value today

[scottmitchell74]

I just "discovered" this part of B-R.

What a fun toy! If nothing else it gets my man 33 over 2000 RBI where he belongs and has Arod below 2000.

Was surprised to see Yaz's stats almost identical: I thought his career pretty well coincided with a down offensive era and not historically almost dead-neutral. ??

[layson27]

I just "discovered" this part of B-R.

What a fun toy! If nothing else it gets my man 33 over 2000 RBI where he belongs and has Arod below 2000.

Was surprised to see Yaz's stats almost identical: I thought his career pretty well coincided with a down offensive era and not historically almost dead-neutral. ??

This matches what my normalized batting line for him looks like as well. I think the park effects of Fenway have a lot to do with his stats looking almost neutral after conversion. Besides, 1967 & '68 were really the only two historically down years of the mini deadball-era (1963-1968)

[scottmitchell74]

This matches what my normalized batting line for him looks like as well. I think the park effects of Fenway have a lot to do with his stats looking almost neutral after conversion. Besides, 1967 & '68 were really the only two historically down years of the mini deadball-era (1963-1968)

Ah! makes sense.