If you take a player of .250 and multiply it by 4, that player would be worth 1 run a game. You take this and multiply it by the number of games the player has played and you will see how much the player might average at the end of the season. Since this player would average a run a game, over a 162 season, the player would score close to 162 runs.
If a player for 162 games had 4 at bats, 1 homerun and 3 strike outs every at game. His PoA would also be worth .250. Obviously, hitting a home run is an automatic run for the player and he would exactly score 162 runs.
Now let's say the player had 1 home run and 1 double every game plus 2 strike outs. The player did this for 162 games.
With the formula the player the player's PoA would be 0.375, or you could say the player is worth 37.5% of a run per bat. since the player averaged 4 bats per game he would be worth 1.5 runs per game. Take this number and multiply it by the games he played and this player would likely score near 243 runs on the season.
Let's take a real world example now, i am going to pick a random player and see if the number's are close. The Player is Dustin Pedroia.
Dustin played in 141 games this year. Dustin had 163 hits, 39 doubles, 3 triples, 15 home runs and 48 walks in 623 plate appearances. Dustin hit 106 sinles that year. Now using the forumla his PoA is 0.1207865. Dustin averaged 4.4184397 plate appearances per game. You take his PoA of 0.1207865 and multiply it by his average plate appearances per game of 4.4184397, you would get 0.5336878. In other words he averaged slightly over half a run per game or he would average just over a double per game.
Take his 0.5336878 and multiply it by the number of games he played and it reads that Dustin should average near 75.249979 runs on the season. Dustin Pedoira scored 81 runs this year. You could say Dustin's poA of 0.1207865 was worth slightly over 12% of a run per bat if you think about it.
Sorry, but this doesn't have any sort of actual worth in terms of estimating value. Your "formula" is flawed and doesn't fill a void in the SABR world.
I don't want to discourage effort, but you aren't getting anywhere with this. The math and reasoning is WAY too basic.
First of all, why do you divide the title by the number of plate appearances and then multiply it by PA per game, when you can just skip that altogether and do BB(.25) + 1B(.25) + 2B(.50) + 3B(.75) + HR (1) and get the same result?
Secondly, if we're translating this to runs, why wouldn't we just use the exact percentages in the formula for BB, 1B, 2B, 3B and HR to make it more accurate? If the number is useful mainly to figure out how many runs a player is worth, I don't see the point of using physical distance as the percentages.
According to what you said you want to accomplish you want to create SLG on a different scale. Which is useless, especially since I already discussed how SLG is also a bad statistic. The idea that you weight a double TWICE as much as a single and so on, is an enormous flaw.
Plus the fact that you are using runs scored as your validation is also hugely flawed, since runs scored is mostly a team statistic and not an individual statistic (well at least its not an individual statistic with any sort of relevance to accurately evaluate the type of player someone is).
Also, what does telling us the physical location of a player on a diamond per game have to do with putting a value (in wins, or runs) of a player?
There is so much that is pointless with your statistic, it just seems like you are reaching for something, but there's nothing there.
wOBA is a wonderful statistic, with lots of scientific research that went into it's creation. Use it, and stop wasting your time with some home made statistic that basically gives us nothing of relevance.
Vampirate, here is the issue with your statistic, which perhaps these other posters haven't clearly conveyed to you. Sabermetric formulas use values for things such as BB, 2B or HR that are based on studies of what their actual values are in terms of scoring runs. Your formula uses values for things such as BB, 2B or HR that you simply made up on a whim, and are in no way related to their actual value. Therefor, your end result is also in no way related to a player's actual value.
Does that make sense to you?
Alright this is your and the majorities opinion and that's fine. I have my own thoughts and obviously people disagree.
However I must say a few things.
1. Any forumula for advanced stats should never change. If you conduct an experiment, you have the ingrediants and you have your formula. When you have your end result you are happy and you believe it is correct. Thus if you find out that just 1 of your ingrediants was wrong, your whole experiment is a failure.
2. A single's value should NEVER change from year to year, neither should a doubles, triples and a home run. those hits values should be static year after year. If they change year after year, then perhaps, just perhaps there's a flaw in the formula.
3. Just because a formula has been researched to death and has been widely accepted does not guarantee it is correct, or even close. In fact, one of the biggest "facts" in science has already been debunked.
And yes I can prove point 3 if you wish.
If a Home Run isn't worth 1 run, then what is it worth?
And please, don't just put a forumula up because as I said, even the most accepted forumulas in history could be proven dead wrong and way off base.
If there is just 1 flaw in the formula then the whole equasion is useless.
Do not just give me a forumula, EXPLAIN how the formula makes sense.
Give me an example.
If there is one flaw in a stat, then it failed miserably? Seriously?
Then why are you even trying to come with a formula in the first place? No stat is ever going to be 100% accurate.
I don't get why the value of a single, etc. can't fluctuate either. A player's value is compared to the productivity of the rest of the league. The league fluctuates. 10 years ago, this was an offensive force. Now, the offense is greatly down. Surely that means that the value of a single, a double, a triple and a home run is going to fluctuate. You compare a player to each season.
Last edited by Driven; 10-24-2012 at 06:52 PM.
You've created a formula based on how far a man runs based on a plate appearance. I could point out how you've completely ignored walks, but even still, who cares? What does your stat actually tell us, or attempt to?
A stat like wOBA is supposed to tell us how good a player's offensive performance is, regardless of the team around him. Mainly, what do various types of hits and walks mean in regards to actually scoring runs, based on what has happened. So what is a double truly worth? Well, we can look at every example of a double from last season, and every other season before, and determine how many doubles led to runs scored. And was it one run? 2? 3? More???? And we can do that for every single, double, triple, HBP, etc that transpired. And using those numbers we can see how valuable a double is, based on everything that has happened in the game thus far. And using those values for every type of hit, we create a formula on a scale fans are accustomed to.
Is it perfect? Of course not. But it is weighted based on actual events. It's certainly a lot more relevant than someone simply grabbing weights for different types of hits out of the air and assuming they hold some meaning.
You can look up which stats correlate to runs scored better than others. You would find that teams with higher OBP score more runs than those with high batting averages. And you would find that a stat like wOBA tends to be a pretty darn good indication of offense. Teams that have guys with high wOBAs tend to score more runs than those that don't.
Is that always the case? Of course not. Then again, not all those who chain smoke die of lung cancer. And some who have never smoked do. But we know better than to think that cigarettes and lung cancer aren't connected.
Hafner: 44 Games / 8 HRs / .360wOBA
But don't you dare say it ignores walks.
Basiacally my forumula tells what a players bat is worth per plate appearance.
If the player's PoA is 0.250, then the player is worth a single every at bat.
The forumula pretty much tells you at what rate the player got to 1st, second, third and home when put into a forumula.
Once you find out a player's PoA per plate appearance, you just need to find out how many plate appearances the player averaged per game.
Multiply he player's PoA with the average plate appearance per game and you'll see what the player's production would average out per game.
Cabrera averaged a PoA of 0.6816768 per game, or in other words, Cabrera's bat was worth somewhere between a double and a triple a game considering 0.500 would be averaging a double.
Take the player's PoA per game and multiply it by the number of games the player played and it will tell you around how many runs the player should score in that season.
Cabrera PoA for the entire season was 109.74513, he scored 109 runs. pretty damn accurate there.
Of course scoring runs or producing rbi's has factors like how your team does as well. Like I said, the forumula would say around how many runs the player should score that season.
Then there are cases like Andrew Jones, he scored 103 runs. Looking at his batting statistics this season do you believe his bat was worth 103 runs.
The PoA was 0.1294835 per bat, 0.5570987 per game and over the season, his PoA was 90.249989.
This means Andrew Jones SHOULD have scored closer to 90 runs on the season. Andrew Jones' bat was worth 90 runs this season.
Again, the formula doesn't put a value on the hits themselves, it just tells you at what rate did the player reach all 4 bases when added up.
You first find out the player's PoA, then you multiply it by the player's average plate appearances per game, and you'll find out what the player's bat's value is per game. Then you take that number and multiply it by the number of games played and you'll see how many runs the player's bat was worth over an entire season. Usually the PoA of a player for the season is near 5 runs compared to how many runs they actually scored.
You can use PoA to estimate how many runs a player would score in a season.
Go ahead, pick any player from this season and find out how accurate or inaccurate it is compared to how many runs they scored on the season.
You have no idea how excited I am right now.