Do you want to see Miguel Cabrera win MLBs Triple Crown?

[Lex Luthor]

Originally Posted by duncan_idaho

BTW, Most statisticians only count HR once when calculating runs created (Since counting them for RBI and R creates an artificial number and is not statistically sound). Looking at the numbers that way, Cabrera accounts for 74 runs (42+52-19), Trout for 65 (28+49-12). That's not a very big separation in the key stats for Cabrera's case.

The "runs produced" stat has always been the most bogus stat in baseball. It penalizes a player for hitting a home run because he scores a run and drives in a run in the same at-bat. Yet virtually every time a run is scored, SOMEBODY gets credit for scoring it, and SOMEBODY gets credit for driving in the run. That's a total credit for 2 runs produced for every run that is actually scored, EXCEPT when the run scores as a result of a home run.

Adding up RBI and runs scored is a fine indicator of runs produced. Subtracting home runs doesn't give you a better idea of a player's production. All it does is add an element of complexity to a formula that is more accurate when it's kept simple.

[duncan_idaho]

Originally Posted by Brainiac

The "runs produced" stat has always been the most bogus stat in baseball. It penalizes a player for hitting a home run because he scores a run and drives in a run in the same at-bat. Yet virtually every time a run is scored, SOMEBODY gets credit for scoring it, and SOMEBODY gets credit for driving in the run. That's a total credit for 2 runs produced for every run that is actually scored, EXCEPT when the run scores as a result of a home run.

Adding up RBI and runs scored is a fine indicator of runs produced. Subtracting home runs doesn't give you a better idea of a player's production. All it does is add an element of complexity to a formula that is more accurate when it's kept simple.

It doesn't penalize the player. It just prevents counting the same run twice.

When you start combining counting statistics, you have to account for duplication like that.

You want to talk about runs knocked in, sure, you count the HR as an RBI.

You want to talk about runs scored, same thing.

When you want to talk about the runs that Player A contributed to his team in a given period compared to the runs that Player B contributed, though... Player A's home runs still only accounted for one run. Giving him credit for them in R and RBI (without subtracting the HR count from the total) makes it look like Player A contributed (number equal to home runs) more runs than he actually did.

[Lex Luthor]

Originally Posted by duncan_idaho

It doesn't penalize the player. It just prevents counting the same run twice.

When you start combining counting statistics, you have to account for duplication like that.

You want to talk about runs knocked in, sure, you count the HR as an RBI.

You want to talk about runs scored, same thing.

When you want to talk about the runs that Player A contributed to his team in a given period compared to the runs that Player B contributed, though... Player A's home runs still only accounted for one run. Giving him credit for them in R and RBI (without subtracting the HR count from the total) makes it look like Player A contributed (number equal to home runs) more runs than he actually did.

Sure it does. If Mike Trout gets a double and Albert Pujols hits a singe and Trout scores, they each get credit for a run produced. How is that worth twice as much as a run produced by a home run?

This just proves my point that when you try to make statistics too complex, they lose their validity. And regarding WAR, I'm glad you mentioned that Fangraphs and Baseball America can't even manage to agree how WAR should be calculated. That's another thing about WAR that has always made me question its usefulness.

[duncan_idaho]

Originally Posted by Brainiac

Sure it does. If Mike Trout gets a double and Albert Pujols hits a singe and Trout scores, they each get credit for a run produced. How is that worth twice as much as a run produced by a home run?

This just proves my point that when you try to make statistics too complex, they lose their validity. And regarding WAR, I'm glad you mentioned that Fangraphs and Baseball America can't even manage to agree how WAR should be calculated. That's another thing about WAR that has always made me question its usefulness.

I didn't get beyond stat 101 and am not a hardcore SABRE guy. So my explanation is not as complex as a true statistician would make it...

But there is a statistical flaw in counting the same occurrence in two lists, then adding the two lists together without accounting for the double-dip.

Think about the way accounting works. It would be like adding a sale to an individual counter for the salesman, adding it to the total company sales, and then adding the salesman's figures to the total company figures again.