I loooooove winning!

Friday, August 6, 2010

The Art of WAR 2.0 - Hitters Version

Now that we have the hitting values out of the way, we can focus on how to get the fielding values that players bring to the table. Again, in a run form, a run saved is just as important as a run created. In real life, they use UZR (Ultimate Zone Rating) as the basis for the fielding values. Obviously in whatif we do not have UZR or any advanced fielding metrics so I had to get a little creative on how to value a player's defensive value.

2. Fielding Value - The true value of a player in the field is how many runs he has saved your team throughout the year. Given that there is no way of knowing this in whatif, I tried to replicate as close as possible to give a true defensive value of a player. What I came up with as a formula is as follows:

(Plus Plays - Minus Plays) / Total Chances * Game Played

The +/- defensive system that whatif has is the closest thing that I could think of in terms of saving runs defensively. If you make a plus play that either shows that you saved a run, made a play on a ball that would have otherwise been a hit, or did something defensively that provided value to your team. A minus play would be the complete opposite. By subtracting the minus plays from the plus plays you will get your "net defensive benefit" of that player. Did he make more better plays than bad? Next, to see how frequently this happens you must divide this net benefit by the total chances (errors + assists + putouts) of that player to get how frequently he can provide value to your team. For instance, a player who makes 17 plus plays in 200 chances should be a more valuable player than someone who makes 17 plus plays in 600 chances. So, by dividing by total chances you get the frequency of that defensive benefit. Finally, I multiplied this number by the number of games played to add a durability feature to the value. Again, obviously someone who provides a high defensive benefit that plays in a lot of games is more valuable than a player who provides the same defensive value but only plays 60 games a year. You get the gist. Based on Season 16 statistics, the fielder that provided the most value to his team was Bruce Garland of Montgomery. When looking at his stats you could see why. He made 17 plus plays, 0 minus plays and did that in only 280 chances. He was durable as well in playing in 147 games. So his formula looks like this:

(17-0)/280*147 = 8.9 runs saved

**Note - there is no quantifiable way to value catchers defensively and how many runs they save. You can use stolen base % but that could be as dependent on factors outside of the catcher's control that it would not be a useful measure. WAR assumes all catchers are average defensively and credits them with 0 runs of defensive value.

The next step is how to come up with a Replacement Value.
3. Replacement Value - In the case of the Garde Imperial, each game that Clinton Clifton plays, that means nhollan does not have to play a replacement player (defined as totally replaceable, AAAA guy). The more Clifton plays the more time that replacement player stays on the bench, the more valuable Clifton is. The opposite is true as well, the less Clifton plays, the more the replacement player must play, the less valuable Clifton is. So, thanks to the saber guys again they have come up with the analysis that shows that a team full of replacement players would win 30% of the time. They would roughly go 48-114 in a 162 game season. A league average team would go 81-81. The difference is 33 wins or roughly 2 wins per player (9 position players + 5 SP + 3 RP = 17; they don't count the other guys because as bench players they are basically replacement players; 33 / 17 = ~2 wins per player). Based on their analysis they show that those two wins are worth 20 runs, so the average replacement player would COST his team 20 runs per 600 plate appearances. Well, we already established that Clifton's bat was worth 83 runs in season 16. So, now we must also take into consideration that he also SAVED his team a certain amount of runs by not having a replacement player take those atbats.
Calculation is: (20/600) * Plate Appearances = Replacement Value
So, in Clifton's case his formula is: (20/600) * 661 = 22.03 runs of additional value. Basically Clifton created 22 more runs by suiting up and playing those 661 plate appearances instead of having John Smith in AAA come up and take those atbats for the Garde Imperial. This is crucial because it brings durability to the table. If you have someone who can play everyday like Clifton he is way more valuable than someone who can only play 3/4 the season.

So far we have covered hitting value, fielding value, and replacement value. Next post I will cover the positional value and how to convert those factors into WINS. So far you can see how Clifton is the #1 WAR player in 1530. He hits a ton, is durable, and plays everyday thus maximizing his replacement level. Til next time...

No comments:

Post a Comment