It seems to me that this Grich/Clemente debate will never end so long as we expect statistics to tell us which player is better.

What does it mean to be a "better" player? In general, a batter's job is to create runs and not create outs. A fielder's job is to convert batted balls into outs while minimizing baserunner advancement, especially advancement to home plate. A pitcher's job is to prevent runs from scoring.

We can probably agree that a player is "better" than another if he creates more runs in fewer outs. A fielder is "better" if his fielding produces more outs, fewer hits, fewer bases advanced, and fewer advances to home. A pitcher is "better" if he produces more outs and fewer runs scored.

But most of the time, none of these conditions hold completely.

1. If player A creates 70 runs in 200 outs, and player B creates 55 runs in 150 outs, which player is "better"?
   
2. If player C creates 6 runs per game for 5 years, but player D creates 5.3 runs per game for 10 years, which player is "better"?
   
3. If pitcher E has an ERA of 4.00, but pitcher F has an ERA of 1.20 against lefties and 6.00 against righties, which pitcher is "better"?
   
4. If player G creates 4.5 runs per game but is a second baseman, and player H creates 5 runs per game but can only play the outfield, which player is "better"?
   

Without a definition of "better", there's no way to answer these questions. It's not a matter of needing a better statistic, or needing more data, or just getting that other blockhead to see the obvious. It's a matter of defining what "better" means.

We can compute how many runs score from a player's batting line. We can compute how many wins are expected for a team given its runs scored and runs allowed. We can calculate a player's Hall of Fame chances from his career record. But we can never calculate the amount of "goodness" or "betterness" in a player's batting line. There's just no such thing.

The TPR [Total Player Rating] statistic does not measure "betterness". It is not a "rating" of a player's "goodness". It measures an objective fact -- the number of runs a player contributed, above or below the average for his position. If it does not measure that accurately, it is a reasonable criticism of TPR to say so. But it is not a reasonable criticism to say that TPR does not measure which player is "better" any more than it's reasonable to say that it doesn't measure which player is "blorgier."

We can try to define "better," but we all use "better" in different situations to define different things:

1. The player who has contributed more runs above average than another is "better".
   
2. The player who has contributed more runs above replacement for his position is "better".
   
3. If a team of 9 player As would outscore a team of 9 player Bs, player A is "better".
   
4. If adding player A to an average team would contribute more runs than adding player B to an average team, player A is better.
   
5. If substituting player A for the regular at his position would add more runs than substituting player B for the regular at his position, player A is better.
   
6. In retrospect, if player A added more wins to his teams for his career than player B, player A is better.
   
7. If, ignoring any obviously horrible years at the end of their careers, player A added more wins to teams for his career than player B, player A is better.
   
8. If player A is harder to replace at his position with equal batting skill than player B at his position, player A is better.
   
9. If a team wanting to win the pennant this year would draft player A before player B, player A is better.
   
10. If a team wanting to finish in contention for the next five years would draft player A before player B, player A is better.
    
11. Probably more that you can think of.
    

For most obvious cases - Mark McGwire versus Mario Mendoza, say - the answers to all ten situations are the same, and we can agree with each other that McGwire is "better", even if we're using different definitions. But when it gets close - like Bobby Grich versus Roberto Clemente - many of the definitions give different answers, and the answers depend on what we're calling "better". If I'm thinking "better" means definition 9, and you think "better" is definition 1, we're never going to get anywhere, because we're using the same word to mean two different things.

By some definitions of "better", Clemente was "better" than Grich. By other definitions, Grich was "better" than Clemente. There is no way to resolve the question of which player was "better" because the question, as stated, is meaningless.

My personal preference would be for everyone to avoid using the word "better" at all, in any discussion that tries to determine an objective result, and instead tell the list exactly what they're trying to measure. Otherwise, we'll be going along in this thread forever.