Another look at replacement level
I mentioned this idea in passing in the Corey Hart extension thread. In short, I think we're defining 'replacement level' wrong, and the end result is that we're overvaluing guys at 'easy' defensive positions and undervaluing guys at hard ones.
The rationale is fairly involved, but I try to go through it all in a reasonably accessible way.
almost 2 years ago
Jeff Sackmann
15 comments
0 recs |
Email
PrintComments
Very interesting stuff
And it makes a lot of sense. I try to stay away from thinking of players in terms of a specific projected WAR anyways and more of as a range. But I really like what you’re thinking here and it is a bit of an inefficiency. I wonder what could be done to more accurately reflect that in value other than just looking at it in context.
E: George 4 (5, throw, throw, throw, throw).
My main problem with WAR (haven't read your entire post, but don't see this alluded to in any of the headings)
is that it does assume replacement with an, as you say, “amount of production you’d expect from a freely-available replacement—the type of guy you could call up from Triple-A, or grab on waivers, or convince another GM to give you in exchange for a player to be named later.”
however, this does not capture the concept of a better-than-replacement-level-but-lower-than-the-WAR-of-your-current-roster-player. A lot of people when looking at WAR just go “Casey McGehee is 3 WAR” (or whatever he is) without considering that we wouldn’t be replacing him with a 0 WAR player, but rather someone like Gamel, who would probably be about 1.5WAR (again, just pulling numbers out of my hat). So, essentially, in terms of OUR organization, McGehee’s WAR is only 1.5. If you overspend by offering McGehee a $35M/3 yr contract, that’s money that you could have used to upgrade a position like pitching, where we likely are AT replacement level. Say you could offer someone $35M/3 years and have a WAR of 2. It’d be money better spent than the money on the 3 WAR McGehee, as his normative value is only 1.5WAR.
by PagsBrewCrew on Aug 13, 2010 5:56 AM CDT reply actions 1 recs
By your reasoning then
you would have to recalibrate Mat Gamel’s WAR number by the next best available 3rd baseman in the Brewers system, and that player’s WAR number by the 4th best, and so on.
The reason WAR compares all players to a replacement level baseline is to avoid this problem and look at each player in a vacuum.
No one is saying you should blindly look at a projected WAR when evaluating a player without taking into consideration roster construction of that team.
It would be interesting to create a team specific WAR taking roster construction into account, but in that case shouldn’t you also then be taking into account what the market rate for wins is for that team as well?
I don't think you'd adjust gamel's by the 4th best in-house, etc
HE is the replacement, so you say his WARIS* is 0. If he goes down to injury or trade, then you need to adjust projections accordingly, so yes, there might be a need for some sort of “depth” measure (as in McGehee has a 1.5 WARIS [positionally adjusted of course] but a 1.5 there is more valuable than a 1.5 at pitcher). Then when you get down to league-wide WAR, assume that you can swap 0 WAR position player for 0 WAR pitcher, etc, even thought that’s likely not true – the more desperate team will have to give up more than the team that’s just looking for a bit of depth.
but in that case shouldn’t you also then be taking into account what the market rate for wins is for that team as well?
One thing at a time;) Within a team, you should pay approximately the same for equal WARIS. The problem is, on the free agent market, some team WILL need to replace a 0 WAR player and so would be paying them at their WAR level, and not your favorite team’s WARIS level. I think this is part of the reason free agents go for a premium. The teams with the low baseline WAR (coupled with teams that can afford to pay a high premium per WARIS) drive up the price for everyone else to the point that you’re paying at WAR-based levels anyway. But at least for arbitration-eligible pre-FA players, teams should pay based on WARIS, not WAR.
*WARIS = WAR in-system; Made up for this discussion, so if this acronym already means something and I’m using it inappropriately, I apologize.
by PagsBrewCrew on Aug 13, 2010 3:01 PM CDT up reply actions
That would be an interesting stat (especially for analyzing the value players have to their teams) but I'm not sure how it's a criticism of WAR itself
If we assume that WAR is meant to evaluate how well a player performed (replacement level being standardized so you can compare any two players’ WARs).
Ryan Braun: He loves it.
So is it safe to say that WAR is an effective way of comparing players, out of context, from various organizations
But it really isn’t totally appropriate to use when trying to construct a roster…
http://www.mlbsoup.com
Definately true
Not totally appropriate to use without adjusting when constructing a roster because WAR takes into account playing time. If a player is traded to a team where he gets much less, or no playing time, his raw WAR number isn’t really relavent in that situation.
Also, raw WAR is a palindrome, which is always a crowd pleaser.
by Alex Statza on Aug 13, 2010 10:36 AM CDT up reply actions
!~WOW~!
Maybe he should play first base instead. That is, he should lie out there and we can step on him when we get a hit.
Yeah, not a surprise to us, here.
We all know that the surplus trade evaluator is the only tool we need.
http://www.mlbsoup.com
It can be useful in some situations
Especially for projection. It’s an all-inclusive metric that can help you determine which of two options will be better given a chance to play as much as they can (while factoring in injury risk).
E: George 4 (5, throw, throw, throw, throw).
It seems to me like it would be least useful for projection
Given the yearly fluctuations in RAR totals, especially defense. More of a retrospective stat to compare how players performed in a given time period.
Ryan Braun: He loves it.
I disagree
It is good for what you described. But. If you’re looking at your roster for the next year, you’re going to formulate an idea about how well they wiil hit and how well they defend. As well as how many plate appearances of playing time they will get. When you put that all together you’re going to have a good idea of how many wins your team will be projected to get based on WAR (when you factor in pitchers and leverage of course). The strategy has been used before and it’s proved to be very effective in estimating wins.
E: George 4 (5, throw, throw, throw, throw).
I suppose
But aren’t wOBA and career or last-three-years UZR/150 better for that? Granted, those are directly incorporated into WAR year-by-year, but it seems to me like even elite players can fluctuate 2+ WAR pretty randomly from year to year. Adding a team’s component WAR projections to project their record just seems to me like it would be wildly ineffective, but I’ll take your word for it.
Ryan Braun: He loves it.
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Recent FanPosts
NL Central Standings
| W | L | PCT | GB | STRK | |
|---|---|---|---|---|---|
Cincinnati
|
25 | 20 | .555 | 0 | Lost 1 |
|
St. Louis
|
25 | 21 | .543 | 0.5 | Lost 2 |
|
Houston
|
22 | 23 | .488 | 3 | Won 4 |
|
Pittsburgh
|
21 | 24 | .466 | 4 | Won 1 |
|
Milwaukee
|
19 | 26 | .422 | 6 | Won 2 |
|
Chicago
|
15 | 30 | .333 | 10 | Lost 10 |
(updated 5.26.2012 at 6:00 PM CDT)
FanShots
Quick hits of video, photos, quotes, chats, links and lists that you find around the web.
Recent FanShots
Baseball Nation Recent Stories
In Association With
Big League Stew - MLB - Yahoo! Sports
- College player makes impossible over the fence catch to save tournament game (Video)
- Security!: Camden Yards guard collides with Royals’ Mike Moustakas
- The Juice: Nelson Cruz matches career high eight RBIs, Pujols stays hot in Seattle
- Benches clear at Fenway as Red Sox-Rays rivalry heats up
- Keeping good company: Albert Pujols becomes fourth-youngest player to hit 450 home runs
