When we look at fielding at any position we must differentiate between terms like greatness and value. Even the most discerning fans often interchange those terms as if they mean the same thing. Greatness is an esoteric term that often gets debated at sports bars and on television shows on the MLB Network and ESPN. We can look at a number like the index because all three platforms use like terms. Unfortunately, the fielding numbers compare players with the replacement level player and the average player. Putting those together is problematic.
Even the same platform can treat players differently depending on the metric being used. For instance, total zones appears in multiple platforms, but is primarily used by baseball-reference. It compares players to the average fielder at their position. It makes it easy to understand with zero being average. Their dWAR statistic is intriguing to say the least. It compares players with the replacement level performer, but it doesn’t compare them with the replacement level first baseman. It compares them with an overall replacement level player and since first base is the least valuable defensive position, most first basemen are automatically worse than a replacement level performer at another position.
So, we treat the data like ordinal data. The idea is to see if the various sources agree on who the best fielder is and who the worst fielder is. So, we cannot combine the numbers and get any real idea from them. We just show them all to see if the various sources agree on the value of the fielder.
We include the innings because the difference between some of these players in terms of productivity is tremendous. The win shares per 1000 innings has a way of evening those numbers out. This is particularly true when looking at the value numbers. Win shares compares with the replacement level first baseman (unlike total zone runs). Before we look at the year by year data we can get a better idea of where these guys are if we look at the rankings in each category. If we can get universal agreement then we have a good idea of what value each player’s fielding adds to their overall value.
Here is where the rubber meets the road. Recently, I got into a long drawn out discussion on social media given the need for complex metrics and data. As the usual argument goes, we don’t need complex statistics to tell us who the best player is. Of course, this is true to a certain extent, but it ignores a few things and is certainly not true all the time. More often than not, it is simply an overreaction to developments either we don’t understand or don’t have much need for. We see it in other walks of life and all of us are guilty of that thinking to one extent or another.
Those that argue against the use of statistics often use statistics to argue their point. They use batting average, home runs, RBI, and runs scored. They use fielding percentage, errors, putouts, and assists in fielding. It’s incredibly ironic for anyone to use statistics to argue against the use of statistics. Now, we certainly can argue as to which statistics are the most descriptive and accurate, but we should at least be honest about what we are talking about.
That being said, we don’t need statistics to tell us Jeff Bagwell was a better defensive first baseman than Frank Thomas. You could look at them with a glove in their hand and decipher that much. However, looking at them doesn’t answer the question of how much better one is than the other. What is the relative or precise value of Bagwell’s glove in comparison with Thomas? How does that interact with their overall value? Then, we get to more complex questions like how fielding value compares at positions like first base versus say a shortstop or catcher. Is a team better off with a Thomas type or should they go after a Keith Hernandez type?
Whether this kind of data interests the common fan is certainly debatable. No one really needs to know what someone’s secondary average is to really enjoy the game. However, teams increasingly need to be on the leading edge on data to put together the best rosters and determine how much each player should be paid and whether they should be invested in long-term. So, those of us that enjoy looking at data enjoy looking at it because we want to know not only who was the best, but how much better were than they someone else.
When we look at awards breakdowns like these we usually see a discrepancy between the total zone and win share awards and the Gold Gloves. Gold Gloves are voted in by the coaches. Rafael Palmeiro won once after a season where he played over 100 games at DH. This isn’t to say that the coaches are always wrong, but given the fact that they may see an opposing fielder a maximum of 19 times during a season, we can’t really take their suggestions all that seriously. It’s not that they can’t identify a good fielder. They know more about baseball than most of us. The problem is a lack of evidence. So, their voting is largely based on reputations that may or may not be deserved.
The problem here is value at first base is overwhelmingly proportioned on the offensive end. The best fielders are rarely ever the most valuable unless they also happen to be great hitters. We don’t see that very often because the skills needed to be a good fielder (quickness, speed) are not present in power hitters. There are notable exceptions, but if you think of the best fielding first basemen in history they usually did not hit for prodigious power. So, while some of us find fielding fascinating and love to see how players are rated with the glove, in terms of value it is not particularly important in finding out which first baseman was the best in history.