Like us on Facebook


Follow us on Twitter





Page 1 of 3 123 LastLast
Results 1 to 15 of 39

Thread: Calculating WAR

  1. #1
    Join Date
    Jul 2008
    Posts
    5,590
    vCash
    1500

    Calculating WAR

    First of all, here's a really good introduction to wins above replacement:

    http://ussmariner.com/2008/02/14/2008-win-values/

    You can also google "how to calculate WAR tom tango". Tango was the one who first posted this.

    I've posted this at a couple of forums where I go by Hyltzn, so if you've happened to seen this elsewhere, I'm the same person. I'll just copy and paste what I've already posted, except I'll update the spreadsheets:

    I figured since I already had spreadsheet downloaded, I might as well use it for non-school reasons......so I made a Wins Above Replacement spreadsheet. It can be used to find a player's WAR and you can keep it as a database. Anyone can DL and use. I don't give a ****. Anyways....

    Here's the basic terminology:

    Player: the name of the player...

    Position: player's position...

    wOBA: Weighted on base average. You can find the wOBA for every player in the majors here. Just search the player's name.

    BWAA: Batting Wins Above Average. It's wOBA converted into runs and then into wins. I randomly picked this name.

    Fielding: the amount of wins a player is worth on the field. You can use UZR divided by 10.5, or you can eyeball it and use other advanced fielding metrics to guesstimate.

    Baserunning: the amount of wins a player is worth with his baserunning. I don't really know why I included baserunning.

    Adjustment: The amount of wins a player gets added for his position and league. I'll quote Tom Tango here:

    Since the same stats of an average AL player is better than the same stats of an average NL player (i.e., the AL is the better league)
    And the positional adjustment is so a catcher isn't valued the same as a DH or LF. For example, a catcher with a .900 OPS is worth more than a DH with a .900 OPS.

    I should clarify here. Every player gets a replacement-level adjustment based on their league. AL players get +2.5 wins and NL players get +2. Then you give every player a positional adjustment.

    The positional adjustments are:
    +1.0 wins C
    +0.5 SS/CF
    +0.0 2B/3B
    -0.5 LF/RF/PH
    -1.0 1B
    -1.5 DH

    So an AL catcher gets 2.5 + 1 = 3.5

    WAR- wins above replacement

    WAR$- THIS IS ONLY FOR THE 2008 SEASON. It's (WAR * 4.4) + .4

    4.4 = how much MLB teams paid for one marginal win in 2008 FA
    .4 = the league minimum

    Here are the instructions:

    - Go to the spreadsheet of the league you want to look at. Then go to the tab of the player's position. Then fill in the data.

    For example: If you want to look at Pat Burrel's 2007 WAR, you open up the NL spreadsheet and click the LF tab.

    You only fill in the yellow shaded cells. There's player which is the players name. wOBA, which I already posted where you can find that at.

    For fielding, use UZR, though incorporating the fan's scouting report would be a good idea.

    For baserunning, I wouldn't worry too much about this. Dan Fox once popped out the list of the best and worst baserunners using retrosheet data, and the impact they made was very small. Everyone came within 5 to -5 runs, so unless it's someone you know is very good or terrible...don't worry about it.

    Once you fill in the 3 (4 if you include the player's name) cells which you need to fill, the other cells will auto-populate based on your inputs.

    - You have to input the playing time for the player expressed in a percentage. In the previous one I just had the playing time set as .80 (80% of 700 PA). So if a player is projected to or had 560 PA in a season, the playing time would be .8 (because 560 is 80% of 700 PA). For everyday players everyone is usually between the 80% and 90% range, unless it's someone that suffers from a lot of injuries (*coughRichHardencough*). Remember you have to enter the percentage in decimal form.

    - The reason it says .337 or .338 (depending at what league you're looking at) is that I had to go ahead and fill that so I wouldn't get funky numbers (if I leave it blank, it assumes a 0 wOBA, so the spreadsheet looks confusing). Don't worry about that. Just erase the wOBA that's there and fill it in with the player's wOBA.

    - ****Big one here**** There's a difference between true talent level value and one season's value. For example, if you're trying to find out what Mark Teixeira's FA value should be you would have to look at all of his previous seasons and weigh each one differently in terms of importance (because what he did in 2003 doesn't tell you as much as what he did last year). That's basically called regressing to the mean (because his past season is liable to be full of luck or flukey *coughGavinFloyd*). If you're trying to look at what a player's WAR in 2007 was, then you just look at his statistics for that season.


    - Don't worry about park adjusting. I had aWOBA up, but I realized my park adjusting formula, which used the b-r formula, wasn't very good, therefore I just hid it and went back to using regular wOBA. I might incorporate wOBA+ sometime, though.

    Here are the files for hitters:

    AL WAR hitters

    NL hitters WAR



    Pitchers WAR-

    There's the pitcher's version.

    Again, here are the terms:

    Player: Player's name

    IP : Inning pitched


    ERA: Earned run Average

    FIP ERA: FIP plus a league factor so that it fits the scale of ERA. You can find a pitcher's FIP et either Fangraphs.com or THT (you could also use xFIP, which is park adjusted)

    tERA: Just a name I made up so I could balance ERA and FIP. It's just ERA + FIP /2....

    W%: This is not a pitcher's actual W%, it's just a way to express how good a pitcher is. This will autopopulate based on tERA.

    For relievers there is also something called leverage, which, as defined by Fangraphs is:

    A measure of how important a particular situation is in a baseball game depending on the inning, score, outs, and number of players on base, created by Tom Tango.
    You can find Leverage at Fangraphs.com in the player's player page. It's all the way down under the Win Probability section. The one you're looking for is pLI. So if you're going to fill in Mariano Rivera's leverage for the 2007 season, you go here:


    Then scroll down to the Win Probability section, and look for his pLI for the 2007 season, which would be 1.78.

    And if you're using projections (say for the 2008 season or whenever the 2009 projections come out), just go ahead and give closers a 2 leverage, setup men a 1.3, and any other type of reliever a .7. Or you can be more conservative and make it 1.8 for closers, 1.2 for first setup guy, 1.0 for second setup guy, and 0.7 for everyone else.

    08WAR$: A player's marginal $ value for the 2008 season. Remember if you're looking at the win value for a player for any season other than 2008, then you should ignore this column.

    Also remember to differentiate between true talent level and his production for one season. Two different things. I actually made a copy and have 2 different spreadsheets for the 2.

    There's a seperate sheet for closers, and another one for every other type of reliever, since they do have seperate replacement level.

    Here is the spreadsheet for pitchers:

    Pitcher's WAR
    Last edited by Seamhead; 06-22-2009 at 04:46 PM.

  2. 08-25-2008, 04:55 PM
    Reason
    off topic

  3. 08-25-2008, 05:25 PM
    Reason
    off topic

  4. #2
    Join Date
    Jan 2007
    Posts
    9,069
    vCash
    1500
    Excellent, thanks for the heads-up. I've been meaning to ask you for some time now how you calculated yours; I've just been using Colin Wyers' WAR calculator up until now.
    My blog- analysis of the San Francisco Giants, Baseball, and Sabermetrics.

  5. #3
    Join Date
    Sep 2007
    Location
    CO
    Posts
    6,822
    vCash
    1500
    I'm going to bookmark this thread.

  6. #4
    Join Date
    Nov 2007
    Location
    Puerto Rico
    Posts
    30,830
    vCash
    1500
    I don't understand absolutely ANYTHING so I'll just head my way out LoL

    Props.LgnD.

  7. #5
    Join Date
    May 2007
    Location
    Bremerton, WA
    Posts
    6,092
    vCash
    1500
    Quote Originally Posted by PhillyUD26 View Post
    I'm going to bookmark this thread.
    Ditto

  8. #6
    Join Date
    Jul 2007
    Location
    Atlanta
    Posts
    21,725
    vCash
    1500
    No longer!
    Last edited by ugafan; 05-21-2010 at 11:45 AM.

  9. #7
    Join Date
    Mar 2007
    Location
    School in Hoboken, home in Rockland County, NY
    Posts
    54,870
    vCash
    1500
    Very, very good stuff Seamhead.

  10. #8
    Join Date
    Jul 2008
    Posts
    5,590
    vCash
    1500
    Statcorner now has park adjusted wOBA. It's called wOBA*. So you should use that.

  11. #9
    Join Date
    Jan 2007
    Posts
    9,069
    vCash
    1500
    By the way, how are you converting wOBA into wins?
    My blog- analysis of the San Francisco Giants, Baseball, and Sabermetrics.

  12. #10
    Join Date
    Jul 2008
    Posts
    5,590
    vCash
    1500
    Quote Originally Posted by C1Bman88 View Post
    By the way, how are you converting wOBA into wins?
    ((wOBA / lgwOBA)/1.15) * 700 /10.5

    Where 1.15 is wOBA's relation to runs, then you multiply by 700, which is a full season's worth of plate appearances, so that would give you runs above average. Then, you divide by 10.5, which is runs relations to win(and you can fin d that using the pytheganpat), and that gives you wins above average.

  13. #11
    Join Date
    Jan 2007
    Posts
    9,069
    vCash
    1500
    Excellent, thank you.
    My blog- analysis of the San Francisco Giants, Baseball, and Sabermetrics.

  14. #12
    Join Date
    Jul 2008
    Posts
    5,590
    vCash
    1500
    I'm going to release some new versions that will also give you what a player should be making for the next 7-8 years based on the 10% inflation rate that's projected, and the -.5 standard decline rate. You can change a player's decline by simpling editing the equation.

    I'm also using tRA now for pitchers.

  15. #13
    Join Date
    Apr 2008
    Location
    /bin/zsh
    Posts
    3,560
    vCash
    1500
    Great post.

  16. #14
    Join Date
    Jul 2008
    Posts
    5,590
    vCash
    1500
    Alright, these should be a nice tool for this off-season. Once again, remember you have to look at true talent level, and one season, or even two seasons, do not represent true talent level. The minimum I'd use is 3 years, since that's what Tom Tango uses for Marcels. The best thing you can do is use projections (Rally's fielding projections, which I'd be happy to link, and Marcels, or even Oliver), but since projections aren't yet available (Rally's fielding proj. are), the best thing you can do is take a player's last 3 seasons and weight them in order of recency 70/50/35 (if you do that you divide total by 155, FYI).

    AL Position Players WAR

    NL Position Players WAR

    WAR All Pitchers

    Don't mind all the jibberish. Once you start inputting what is needed, then you'll see the results.

    WAR (x year) $: What a player should be making for that season based on true talent (including decline), inflation, etc.

    3-year total: The total amount of money that a player should be make in a 3-year contract (09, 10, 11).

    5-year total: The total amount of money that a player should be make in a 5-year contract (09, 10, 11, 12, 13).

    8-year total: The total amount of money that a player should be make in a 3-year contract (09, 10, 11, 12, 13, 14, 15, 16).

    Feel free to change whatever you want. If you want to use FIP instead of tRA, or ERA instead of tRA, then PM me. You can't just paste the ERA or FIP since both are on slightly different scales than tRA.

    I have the decline set as -.5 for every season, but you can change that if you'd like. I change mine depending on the player. PM me if you want to know how to change this without messing up, too.
    Last edited by Seamhead; 11-14-2008 at 09:45 PM.

  17. 11-14-2008, 10:24 PM
    Reason
    Off topic

  18. #15
    Join Date
    Jan 2007
    Posts
    9,069
    vCash
    1500
    Thank you very much.
    My blog- analysis of the San Francisco Giants, Baseball, and Sabermetrics.

Page 1 of 3 123 LastLast

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •