In this post I'll start getting into what goes into the numbers for position players, and how their performance translates into their WAR statistic. Again, the great thing about WAR is that it is simple and easy and everything is based on runs, which you know, are kind of important in baseball. The basis of WAR is to find out HOW MANY RUNS OF VALUE a certain player brought to his team in a given season. (ohbytheway, did you know that Jonny Gomes is a 0.0 WAR player so far this season. He is EXACTLY replacement level. Lance, I can't believe you haven't been hammering this home yet. He has performed like a AAAA guy this season yet continues to play everyday and hit 5th. Chris Heisey on the otherhand has produced a 1.5 WAR in roughly 300 fewer plate appearances. Basically if Dusty had given all of Gomes' atbats to Heisey, he would have provided the Reds roughly 4.5 MORE WINS of value than Gomes. Incredible. Okay, back to how to calculate WAR...) So, there are 4 ways that a position player can provide value to his team. They are:
1. Hitting Value - how many runs did he produce with his bat?
2. Fielding Value - how many runs did he save with his glove?
3. Replacement Value - how many runs did he create for us based on the fact that he was able to play and we didn't have to call up some hack from AAA and have him play for us?
4. Positional Value - how many more runs is this guy worth because he can play SS for us and not DH? Obviously it is harder to find a 5.0 WAR SS than it is to find a 5.0 DH. We need to account for that.
We'll dive into these one at a time...
1. Hitting Value - Batting average has been historically used to value how good a player is with the bat. But is it really the best way to value a player? Is Rafael Furcal (.316) creating more value for his team than Albert Pujols (.305)? I'm going to go out on a limb and say "no." OBP% is nice because it measures a players ability to not make outs, which again, in baseball is important, but it doesn't tell me anything about what he does when he doesn't make an out. Did he walk? That's valuable, but not as valuable as say did he hit a triple? SLG% is nice because it quantifies how many bases a player accumulates when he hits the ball, but again, its flaw is that it doesn't measure how often a player gets on base. Again, is Corey Hart (.559) providing more value with the stick than Jayson Werth (.516) even though Werth gets on base at roughly a 40 point clip higher than Hart? Tough to say. That is why we use wOBA as the main driver for the hitting value. Basically, through statistical analysis saber guys have found that OBP% is roughly 2 times more important than SLG% so wOBA = (2*OBP)+SLG/3. It really is a lot more complicated than that but this is the easiest way to explain it. So, for an example, we will use Cliton Clifton again:
(2*.463 OBP%)+.629 SLG%/3 = wOBA of .518
Now all we have to do is compare that to the league average wOBA to find out exactly how much more valuable Clifton was than the average player in 1530 last season. The average wOBA was .367 so Clifton was .518-.367 = .151 more valuable. But, how do we convert that to runs? Simple. Take that number and divide by a constant (1.15) and multiply that by how many plate appearances the player had and you get the amount of runs that Clifton created over the average player last season. So, (.151/1.15)*661 plate appearances = 87.17 runs above average.
Now, that we have that, there is just one last piece of the puzzle that is missing from Clifton's hitting value. The question is, how can those 87 runs be as valuable in Montreal (park effects 2, 2, 2, 1, 1) as in Burlington (park effects -4, -3, -3, -4, -4)? The obvious answer is that they aren't because since runs are harder to come by in Burlington, they are more valuable runs than if they were created in Montreal. So, we just need to adjust for park factors. Since I don't have the time (nor the manpower) to calculate park effects on runs scored in 1530 Homer, I took a simplified process in doing this. Basically add up the park effects and adjust half of the player's value runs by that amount on a percentage basis.
Clifton Park Effects = 8 (2 + 2 + 2 + 1 + 1)
Half of Clifton's Runs (I used half because he played half his games in his home park and I again took the simplified process of all of his road stats would just even themselves out by playing in hitter positive, negative, and neutral parks) = 43.58
Park Effect - Negative 8%*43.58 = -3.46 runs
Adjusted Runs Above Average for Clifton after Park Effects = 87.17 - 3.46 = 83.71 Runs
*(on the flip side, Burlington's hitters would receive a boost of 18% (-4, -3, -3, -4, -4 = -18) to half of their run values because their home park depresses offensive output.)
So, Clinton Clifton provided Montreal with 83.71 Runs Above Average in season 16 with his bat. My next post I'll explain how to come up with the other three factors in a position players' value. Hope this wasn't too complicated or boring, and that you'll be getting some value out of it. I know I enjoyed learning about it. Til next time....