WAR

gehrenberg   7/14/2014 2:20 PM

Here is the article with the equation for wins above replacement http://www.fangraphs.com/library/war/war-position-players/

jyang61107  7/14/2014 2:22 PM

Now here was the debate.

I agree that WAR is valuable with Fantasy and numerous people disagree to it as not.

Open for debate on why and why not.

On a side note I deleted all the comments on the player page in order to make it relevant to that player page and will post here for the debate. 

seahawks126  7/14/2014 2:24 PM

"I agree" implies that someone else made this point and you are agreeing with it. There's no one for you to agree with because no one has made this point.

seahawks126  7/14/2014 2:26 PM

You shouldn't be making an argument about a statistic when you don't even understand how it is calculated.

jyang61107  7/14/2014 2:27 PM

I can agree to myself if I choose to and its up to you to make a valid case for me to sway from that decision.  I'm just trying to understand fantasy to the fullest. I just can't have someone tell me No because its NO. Explain to me why WAR is useless and why I should be considering looking into something else like how gehrenberg ask me to look into other factors.

gehrenberg   7/14/2014 2:28 PM

I think the better question is what about WAR should be used for fantasy. It doesn't take into account any fantasy stats, it is weighted for opponents faced, defense and baserunning are heavy components as well as position played. A hitter who plays in Colorado sees a significant decrease in their WAR because it is a strong hitters park, but for fantasy purposes it would increase value. 

jyang61107  7/14/2014 2:30 PM

Statistic are not always right.  That is why I might fall into the fallacy of statistic.  Care to explain and help my knowledge or just keep the hate.  I'm just asking a question and if you want to burn me feel free too but that doesn't help in any way here for the community.  Maybe some other people here are learning about this too, you know and I happen to have the courage to throw it in here and get flamed at.

jyang61107  7/14/2014 2:31 PM

@gehrenberg. This would make COL hitters more valuable in WAR.  Wouldn't that make them good in batter's categories in the fantasy game?

gehrenberg   7/14/2014 2:34 PM

No it makes COL hitters less valuable in WAR because WOBA adjusts for park played in. WOBA adjusts offensive value in terms of runs created as if all players were playing on a neutral field.

seahawks126  7/14/2014 2:35 PM

Fantasy baseball is about predicting tangible stats that are counted in fantasy (I.E. R, HR, RBI, SB, AVG). WAR has nothing to do with those stats, and instead has to do with a player's ability to create real life wins for a real life baseball team. These things are immensely different and that is why WAR is a poor measure of fantasy value. COL hitters are not more valuable in WAR because R, HR, RBI, SB, AVG have nothing to do with WAR calculations. These stats get inflated by a good ballpark, but WAR doesn't care. It makes an adjustment to neutralize the park factors.

jyang61107  7/14/2014 2:38 PM

@gehrenberg  Therefore we should look at WOBA?

@seahawks126 A team wins by a score of 1+.  Wouldn't a higher WAR team rated most likely to win and if so are more likely to get those batter categories in the games? What should I be looking at? WOBA like gehrenberg advises?

gehrenberg   7/14/2014 2:41 PM

No I am saying not to look at WOBA for fantasy because of the park adjustment. And WOBA and WAR are related because WOBA is the offensive component of WAR. 

seahawks126  7/14/2014 2:43 PM

You shouldn't be looking at anything that has to do with a team winning. You should be looking directly at what metrics have predictive value for fantasy categories. While there is bound to be some correlation between winning teams and fantasy categories, it's entirely unnecessary to try to derive value that way. There are much better metrics available that translate directly to R, RBI, HR, SB, and AVG. wOBA has to do with WAR, which again is missing the point of predicting fantasy value, as opposed to team wins.

gehrenberg   7/14/2014 2:44 PM

For fantasy the most valuable stats to look at are the underlying stats that corellate with the fantasy stats. For example BABIP is directly related to batting average, and swst% for pitchers directly leads to strikeouts. There isn't one all encompassing stat to look at. Rather you want to take several factors into consideration when judging how a players performance will be for the remainder of the season. 

jyang61107  7/14/2014 2:52 PM

Would luck play a factor in BABIP? If so how can we look at BABIP over WAR.

I agree in the averages for batter's and pitcher's because those are real numbers coming out.  

Can't we use WAR and BABIP together to get your ideal candidate of an elite player from average?

gehrenberg   7/14/2014 2:56 PM

BABIP is more of a number that measures the luck of a player. So if a player has a really high BABIP they have been lucky. If the BABIP is really low they have been unlucky. But even so there are factors such as speed and line drives that play into BABIP that can cause it to to be higher or lower than league average. WAR doesn't really have any relation to batting average.

gehrenberg   7/14/2014 2:59 PM

The other issue with WAR for fantasy purposes is that is has no value as a predictive stat it only takes into account past performance and fluctuates greatly from year to year. 

jyang61107  7/14/2014 3:00 PM

If we do stay away from injured player how do we buy the ones that might come out good? The A. Beltre type.

gehrenberg   7/14/2014 3:03 PM

By Chris Davis do you mean a player that gets injured? Injuries are very difficult to predict. But in general stay away from players with injury history and players without a track record of playing at the level they played at the year before. I would expect Chris Davis to be a good fantasy option when healthy. 

jyang61107  7/14/2014 3:06 PM

Agree with the 1 season breakout.

gehrenberg   7/14/2014 3:09 PM

But it is also about value and risk/reward. I wouldn't take Chris Davis in the first round next year. But assuming he is going to be healthy and hitting in the middle of the Orioles lineup he is probably a great value in round 6 or so next year. 

jyang61107  7/14/2014 3:15 PM

Judging from using BABIP any players that catches your eye in 1 buy low and 1 sell high moment?

I will choose to ignore WAR since reading more articles on BABIP where it gets the edge over WAR.

But I still hate the fact when projected good team player uses WAR.

Any chance we can use WAR as a secondary choice if BABIP is really close? and if not what should be next.

gehrenberg   7/14/2014 3:35 PM

I wouldn't focus one or two numbers you should take your time going through fangraphs and learning what all the different statistics mean so you can derive as accurate of opinions on players as possible

jyang61107  7/14/2014 9:57 PM

So far I read into WAR. BABIP XFIP FIP and SIERA. What other do you suggest and thanks. 

descender  7/15/2014 12:35 PM

I disagree with some parts of this discussion. I agree that WAR is not the right stat to use per se, but I believe that some of it's components ARE useful in identifying good players for almost any fantasy league. I love using wRC+ and wOBA to seek out players that are performing well beyond the confines of the 5x5 or 6x6 categories of my leagues. Predicting Runs or RBI's alone are nearly impossible. I believe you are better off acquiring purely productive hitters and not completely focusing on the counting stats. The theory being that Runs/RBIs or Wins/low ERA should follow naturally.

This is the FAQ for the Tango linear weights points system https://docs.google.com/document/pub?id=1i_k-uYkFQDR7M9aomdqdvltDwhye-uI7wSJKtRzPBT4&pli=1.  The FAQ is a little outdated, and the current system now uses hits against for pitchers.

There are links in that FAQ that explain how the batter and pitcher points are derived. Long story short it's basically wRC+ and FIP. I'll include the points here, and I suggest at the very least creating this league and importing your players. Look at it once before you judge me crazy. :)

Punching this into the BBM projections (that accommodates for lineup position et al) creates what I consider to be extremely accurate and stable projections for a players "skill" level.  I find it especially useful for daily/weekly projections, because trying to predict things like Runs/RBI's is extremely inaccurate.

It's a great list to start with, anyway. Only when you have a good list can you use things like BABIP, HR/FB, or batted ball splits to decide if that performance is sustainable or due for regression (in either direction).