Fun with Win Values
by kenshin1 on Apr 28, 2009 11:50 AM PDT
Email
PrintAs many of you know, the win value stat represents the latest, and by far the best IMO method to evaluate players. Using the best available offensive stat (wOBA), defensive stat (UZR) and pitching stat (FIP), win values literally allows one to assess how many wins a player has been worth compared to an equivalent minor league replacement (think Willie Bloomquest). Additionally, since major league teams pay roughly 5 million/ win on the FA market, win values enables us to think of player value in monetary terms (which is ridiculously cool and intuitive). Since MGL generously allowed access of UZR to fangraphs, win values are now updated continuously throughout the season. So how have the Giants done throughout April?
Comments:
wOBA is a great stat regardless of the sample size; however, UZR has a lot of noise this early in the season and should be taken with a grain of salt. There is a positional adjustment which I did not include in my chart (CF defense is worth more than 1st base defense for instance), and which is why adding defensive RAA and offensive RAA does not completely explain win value.
1. Lincecum is unbelievably good. To wit: as good as Cain has been this season, he has only contributed about 1/2 as much as Timmy. Lincecum has earned many times over his actual salary this season.
2. Summation of the Giants 2009 offense: Andres friggin' Torres has been the 3rd best position player this season (depending on how much you believe the defensive stats)
3. Despite his admittedly awful offense, Renteria has been OK so far this season. UZR likes his defense (which jives with my subjective visual evaluation). The dude cannot hit but Worst. Contract. Ever. this ain't.
4. Travis Ishikawa YIKES. Emmanuel Burriss DOUBLE YIKES. If the world was a meritocracy, these two would PAY the Giants 4 million dollars for their play this season. That is how much worse they have been than AAA veterans. FREE KEVIN FRANDSEN.
5. Barry Zito- 3rd best Giants starter this season. <Cue weeping and gnashing of teeth>. I am quite frankly terrified and astonished.
6. I was completely on board with the Randy Johnson signing at the time. In reality, it has not worked out very well so far.
7. According to UZR, Aaron Rowand has no business playing CF. I hate this signing more with every passing day.
|
Player |
Defensive RAA (runs above average) |
Offensive RAA |
Win Value |
Monetary Value (in millions) |
|
Fred Lewis |
2.6 |
3.9 |
.8 |
3.5 |
|
Bengie Molina |
1.6 |
.5 |
2.3 |
|
|
Andres Torres |
2.1 |
1.0 |
.4 |
1.6 |
|
Edgar Renteria |
1.1 |
-2.4 |
.2 |
.8 |
|
Pablo Sandoval |
-1.0 |
-.3 |
.2 |
.7 |
|
Randy Winn |
.5 |
-1.9 |
0 |
.2 |
|
Aaron Rowand |
-2.7 |
.3 |
0 |
-.1 |
|
Rich Aurilia |
.4 |
-2.3 |
-.2 |
-.8 |
|
Eugenio Velez |
-1.3 |
-2.2 |
-.3 |
-1.3 |
|
Travis Ishikawa |
1.1 |
-5.7 |
-.4 |
-1.8 |
|
Emmanuel Burriss |
-.2 |
-7.0 |
-.4 |
-2 |
|
Pitcher |
Win Value |
Money Value |
||
|
Tim Lincecum |
1.1 |
5.1 |
||
|
Matt Cain |
.5 |
2.2 |
||
|
Barry Zito |
.2 |
1 |
||
|
Jonathan Sanchez |
0 |
.1 |
||
|
Randy Johnson |
-.1 |
-.5 |
This FanPost is reader-generated, and it does not necessarily reflect the views of McCovey Chronicles. If the author uses filler to achieve the minimum word requirement, a moderator may edit the FanPost for his or her own amusement.
109 comments
|
3 recs |
Do you like this story?
Comments
Weird that Lewis is our best defensive player.
Merkin Valdez? Manuel Mateo? A rose by any other name...
remember...
that RAA for position. He’s not our best defensive player, he’s just much better at playing LF than most left fielders are.
Flossing a dead horse
it still confounds me. his play seems to have been slightly poor. perhaps i am only remembering the bad, but boy howdy he had some misplays. i guess he has made other plays that more than made up for them…
i adopt Les Plack, D.D.S. for the following reasons:
1. greatest name for occupation ever (except maybe lesbian prostitute sharon bottoms)
2. i am hoping for free dental care from my son (if he ever hears of this)
3. the giants need him more than anyone because their smiles are all they've got right now
by Headhunter Rollins on Apr 28, 2009 11:58 AM PDT up reply actions
yeah he compensates...
for his bizarre routes and terrible jumps by sheer athleticism. And remember, most left fielders are statues who really shouldn’t be playing in the field.
Flossing a dead horse
cough Manny cough Ramirez
Minor White > Ansel Adams
by say hey nation on Apr 28, 2009 12:10 PM PDT up reply actions
he has CF range and DH instincts
Randy Winn is going to catch that. And he'll do it real classy-like too.
I expected Lewis to be on top. Renteria has picked it up in the past week, so it doesn’t surprise me to see him higher than some of our other hitters. Also, you have Aurelia on there, but where’s Lowery?
Judgment Day is coming
comics | art | Nattowear
Terrell Lowery?
Awesome 42 PAs.
Meet my new son: Sundrendy Windster, coming soon to a minor league near you.
by EliminateMe on Apr 28, 2009 12:25 PM PDT up reply actions
So...
Manny Burriss is worth -$2mil. Does that mean we should pay him $2mil to go back to Fresno? Or that he would have to pay $2mil to play on a MLB roster?
Baseball is a lot like life. The line drives are caught, the squibbles go for base hits. It's an unfair game. -Rod Kanehl. Adopted brother of Jason Jarvis
Neifi Perez...
once “owed” the Royals 8 million dollars over a whole season. I believe that’s the worst total since UZR was invented.
Flossing a dead horse
Quick groupthink check: Manny Burriss is actually really good at second base, right? And Rowand has rebounded somewhat in center?
I’ve been pleased with Rowand in center this year, FWIW.
by Grant Brisbee on Apr 28, 2009 12:20 PM PDT up reply actions
maybe we can dress him up like a ninja
Minor White > Ansel Adams
by say hey nation on Apr 28, 2009 12:23 PM PDT up reply actions
I think Rowand has held his own so far this year in center, but he is no Eugenio Velez out there.
As for Burriss, he has the athleticism/tools to be really good at second but has made a few head scratching plays so far. So the jury is still out IMO.
Baseball is a lot like life. The line drives are caught, the squibbles go for base hits. It's an unfair game. -Rod Kanehl. Adopted brother of Jason Jarvis
I just meant defensively
probably should have clarified that one.
Baseball is a lot like life. The line drives are caught, the squibbles go for base hits. It's an unfair game. -Rod Kanehl. Adopted brother of Jason Jarvis
As I posted elsewhere....
…although Renteria started the season ice cold, his line over the last 10 games is .314 /.415 /.571 .
As dismal as Burriss’ batting line looks – it’s still better than Bocock’s from last year. I still have hope for him but he needs a lucky streak so he can stop pressing so hard.
Meet my new son: Sundrendy Windster, coming soon to a minor league near you.
Giants new advertisement:
At least our hitters are better than Bocock!
by paboperfecto on Apr 28, 2009 12:52 PM PDT up reply actions
When you replace one crappy young player with a less crappy young player, we call that “baby steps”
My adopted Giant: "Raptor Jesus" Guzman
Thank you Dr. Marvin
Giant Dirtbags: Brian Anderson, Todd Jennings, Steve Hammond, John Bowker
Wronghanded Affeldt doing it right
by Giant among Angels on Apr 30, 2009 9:09 PM PDT up reply actions
I don't think Burriss has been much better than Bocock...
Bocock 2008 (93 PAs): .143/.156/.258 OPS+ of 12
Burriss 2009 (66 PAs): .169/.186/.246 OPS+ of 13
Sadly, those numbers are pretty similar.
Baseball is a lot like life. The line drives are caught, the squibbles go for base hits. It's an unfair game. -Rod Kanehl. Adopted brother of Jason Jarvis
Man, those are some hilarious lines.
Sometimes, nowadays, I start feeling pretty good about Giants management. Then I remember the silliness with Bocock & Burriss over the last couple of years, and I have to conclude that Brian Sabean and his crew are barely qualified to manage a frozen-banana stand.
Frozen Banana stand
If this doesn’t call for a Sabes and Baer strip, I don’t know what does.
"Catcher are base running. Hitters are offense."
Only [hella] games left until the end of Zito's contract.
Adoptive father of "Poncho" Villalona: This Angel don't fly. Nothing about him is light.
TWSS!
Giants pitching coach Dave Righetti. "I treat Timmy differently from most pitchers: I leave him alone."
card carrying Bochy Hatter
by natteringnabob on Apr 30, 2009 10:03 PM PDT up reply actions
Caption
“I have a giant banana. Do something about it.”
by Lars The Wanderer on Apr 29, 2009 2:14 PM PDT up reply actions
I didn't say MUCH better.
I said better. And I’m holding onto that small thread of comfort, thank you very much!
I really believe Burriss is better than this, he’s just let the pressure get to him. I’m glad the Giants are showing patience with him. But that’ll only last so long.
Meet my new son: Sundrendy Windster, coming soon to a minor league near you.
RBIs are for selfish stat padders. The real metric is game winning RBIs.
When I was a kid I used to pray every night for a new bicycle. Then I realized God doesn’t work that way, so I stole one and prayed for forgiveness. - Emo Philips
Proud father of Juan Carlos Perez. Think Albert Pujols at second.
I bet it doesn’t even factor how a pitcher should will it’s team to score runs.
by Lars The Wanderer on Apr 28, 2009 1:10 PM PDT up reply actions
Interestingly enough...
Matt Cain has a -4 run positional adjustment for “not being a winner”
Flossing a dead horse
you're just jealous...
that someone else writes nerdy things about the Giants.
Flossing a dead horse
Spreadsheets aren’t nerdy enough…
Aaron King is still my homeboy... iffy mechanics and all
If Dustin Pedroia played in Seattle, not many people would be talking about him.
So Panda is at -1 defensively, do you think he can continue to get better or has he been lucky in a small sample size
The absentee father of Edwin "Cue Ball" Quirarte, now that he is doin well in the minors, its time to get back in his life, and in his checking account!
you have to multiply that out over a whole season....
assuming -1 represents Pablo’s true performance and he continues at that level over an entire season, his UZR would be -8.3/150. If that was the case, one could pretty accurately describe him as a -10<x<-5 third baseman (IE: one who can barely play the position)
Flossing a dead horse
Damn the anecdotal evidence in me that says I like him at the hot corner
The absentee father of Edwin "Cue Ball" Quirarte, now that he is doin well in the minors, its time to get back in his life, and in his checking account!
I still feel like the hitting is bound to improve in parts – Ishikawa and Burriss might not be good players but they aren’t mid-teens OPS+ players. In particular, it seems like Travis has been the victim of some poor luck at times.Unless he’s suddenly fallen off a cliff, Randy isn’t a 75 OPS+ player. Even though I like to make fun of him, Rich probably isn’t quite this bad either.
Then again, I suppose Fred probably isn’t really a 133 OPS+ player and Bengie isn’t really a 117 OPS+ player. BUT the bad numbers I think are bound to go up more than those two are bound to go down.
It’s not a good offense, but I really don’t think it’s going to be quite this bad all year. But maybe I’m just feeling optimistic.
btw, I <3 Fred Lewis
Matt Cain: He'll save children, but not the Dodger children.
PABLO SANDOVAL AM STEAL DEATH, DESTROYER OF WORLDS.
Fred Lewis: 133 OPS+
Manny Burriss: 13 OPS+
Does this mean that a lineup of nine Manny Burrisses would not out hit Fred Lewis?
Baseball is a lot like life. The line drives are caught, the squibbles go for base hits. It's an unfair game. -Rod Kanehl. Adopted brother of Jason Jarvis
a lineup with one Barry Zito does owe us $18 million.
Proud adopted parent of future big league slugger Thomas Neal
Official Sponsor of the 1997 San Francisco Giants
by nostocksjustbonds on Apr 28, 2009 1:44 PM PDT up reply actions
This
this thisity this
Giants pitching coach Dave Righetti. "I treat Timmy differently from most pitchers: I leave him alone."
card carrying Bochy Hatter
by natteringnabob on Apr 30, 2009 10:04 PM PDT up reply actions
Now that’s what I call a win Winn…situation.
by Bay Area Sports Guy on Apr 28, 2009 1:42 PM PDT reply actions
Win Wynn?
I only have a signature because I recognize everyone else by their sigs, not their usernames..
I Too Am A Fan Of Win Values
But I have noticed that their are still many faults or at the very least quirks with how the various elements are calculated. One very recent example I can think of is that yesterday’s game resulted in Aaron Rowand’s UZR going from -1.7 to -2.7. I watched the game and I can’t figure out how this could be explained. Can anyone tell me just which ball hit to centerfield Rowand didn’t get to that he should have or just what throw Aaron didn’t make that he should have?
The bigest fault I see is that UZR and FIP do not agree consistently enough with each other which to my way of thinking shows that they both still contain many errors in how the measure defense and pitching. What I would expect is that a team that has a +UZR would have and ERA lower then their FIP while a team that has a -UZR would have and ERA higher then their FIP. There are just too many examples where these are not the case.
One very recent example I can think of is that yesterday’s game resulted in Aaron Rowand’s UZR going from -1.7 to -2.7. I watched the game and I can’t figure out how this could be explained. Can anyone tell me just which ball hit to centerfield Rowand didn’t get to that he should have or just what throw Aaron didn’t make that he should have?
UZR is compared to average. It’s possible that there were just some good plays by CF around the league.
What I would expect is that a team that has a +UZR would have and ERA lower then their FIP while a team that has a -UZR would have and ERA higher then their FIP. There are just too many examples where these are not the case.
Looking at the numbers, there seems to be a pretty clear correlation between UZR and ERA-FIP. Look at the top and bottom five in UZR, and how they ranked in ERA-FIP:
Top 5 in UZR:
- Phillies. ERA-FIP: 3rd best.
- Rays. E-F: 2nd best.
- Athletics. E-F: 8th best.
- Astros. E-F: 9th best.
- Red Sox. E-F: 12th best.
Bottom 5 in UZR:
- Padres. ERA-FIP: 18th best.
- Tigers. E-F: 21st best.
- Rangers. E-F: 30th best.
- Dodgers E-F: 15th best.
- Rockies. E-F: 19th best.
The top 5 teams in UZR all had an ERA lower than their FIP. The bottom five in UZR all had an ERA higher than their FIP.
Obviously there isn’t a perfect correlation, but there’s no reason to expect one. Think about this example:
Pitcher A allows a HR, then walks two batters, then strikes out the next three. His ERA is now 9.00. His FIP is now 16.2. ERA – FIP: -7.2.
Pitcher B walks two batters, then gives up a HR, then strikes out the next three. ERA: 27.00. FIP: 16.2.
ERA – FIP: 10.8.
Pitcher A has a way, way lower ERA – FIP, but that has nothing to do with the defense behind him – there were no balls in play. This is just an example for one of the explanations for a substantial differnce between a pitcher’s ERA and FIP, that have nothing to do with defense. Most of those explanation fall somewhere between pitcher mentality and random chance, usually closer to the latter.
Adoptive parent of Noah Lowry. Because he was awesome once, and, goddammit, he shall be awesome once again!
I hope.
Just a thought
I can see what you mean here. It is indeed disconcerting. But perhaps a way to look at it is that had Aaron gone 5-for-5, his average might have jumped from .300 to .350. Did he truly become a .350 hitter? Clearly not.
Statistical measures can get tilted with a small sample. If they are reasonably accurate over a full season, that is what is important.
How exactly
is Torres worth .4 wins with 12 PAs and 21 Innings in the OF.
That’s UNPOSSIBLE.
FIRE BRIAN SABEAN... UNLESS HE KEEPS DRAFTING WELL. .. AND SIGNS UNDERRATED PLAYERS LIKE AFFELDT OR PHELPS. .. OR ALRIGHT WHO'S PLAYING WITH THE ALIEN MIND-SWITCHING RAY?
-------
PARPG- Indy post-apocalyptic roleplaying game currently in early planning stages.
DINGERZ
TATERS
BOMBS
SHOTS
BIG FLIES
GROUGTHINK ALERT
Chatterbalks dot com: Still with jokes. Now with updates.
obviously...
the defensive stats are meaningless in his case. Still it remains a sad testament to the Giants offense that a player with 12 PA’s has been significantly more valuable than many of its starters.
Flossing a dead horse
And what about our super-secret weapon, Nate the Great? Surely he’d rate near the top of our offensive list, wouldn’t he?
Sergio Romo: striking out professional hitters since 2005. And winner of the 2012 NL Fireman of the Year Award!
Secret weapon? You mean legendary creature
Giants pitching coach Dave Righetti. "I treat Timmy differently from most pitchers: I leave him alone."
card carrying Bochy Hatter
by natteringnabob on Apr 30, 2009 10:10 PM PDT up reply actions
Randy Johnson’s value should come up as his HR rate normalizes. It’s something absurd, and it’s going to come down to league average, if not better, given the park we play in. His K/BB rates (besides last start) have been very positive.
Proud father of Barry Zito. As long as he keeps throwing strikes, that is.
The Franchise
really is. The franchise, that is. I likes me some Matt Cain but when you have to stack up against a demi-god you gets some serious under-appreciation! Where would we be without Tim?
Todos somos Gigantes
Cain = Apollo (or am I getting my mythologies confused?)
Hi, I am Johnny Disaster.
by Johnny Disaster on Apr 29, 2009 10:51 AM PDT up reply actions
Cain = Achilles. Just don’t have stand up stuff.
The Basil Fawlty Moderating Strategy:
"We could run a nice blog here if we didn't have all these members getting in the way."
Kevin Frandsen should be with the big team.
Lewis = Odysseus
He spends a long time wandering around places he doesn’t want to be, but eventually gets to the right spot.
Meet my new son: Sundrendy Windster, coming soon to a minor league near you.
by EliminateMe on Apr 30, 2009 11:58 AM PDT up reply actions
Someone explain to me how FIP is the best pitching statistic for this purpose. Please. It doesn’t make any sense. It’s essentially saying: strikeouts, walks, and home runs are the ONLY things that matter for a pitcher. Now, I have an extremely good grasp of statistics, and I have to disagree with this very, very strongly. I understand the value of simplified statistics for forecasting purposes – generally a KISS type method performs well for predicting the future, but in terms of looking back at the past there is SO much more we know to more accurately assess a pitcher’s contributions.
Things like GB/FB ratio, BABIP can tell us a lot about a pitchers past contributions. Even a statistic like BABIP that is not necessarily a “skill”, that is it is not something a pitcher controls going forward, has to mean something to past performance, right? Even if a .200 BABIP is unsustainable, the pitcher DID accomplish it, and it helped the team win – yet we don’t give them any credit for this?
FIP does not measure what a pitcher did.
I.e, you cannot get any clue of “value” – for that just use RA (modified by runners stranded / allowed to score by relievers).
FIP is used because it’s an empirically BETTER predictor of future success than taking account of all the balls in play.
It’s possible that a more complicated predictor that uses GB/FB/LD + BABIP rates exists, but I am not sure it has better results than FIP.
I think the problem with including balls in play is that it completely munges fielder defense into the equation, and the noise added outweighs the value. GRMs point above about UZR+FIP I think points out the flaws and holes in a stat like UZR (which cannot readily account for the “difficulty” of a play made in- or out of- zone)
FIRE BRIAN SABEAN... UNLESS HE KEEPS DRAFTING WELL. .. AND SIGNS UNDERRATED PLAYERS LIKE AFFELDT OR PHELPS. .. OR ALRIGHT WHO'S PLAYING WITH THE ALIEN MIND-SWITCHING RAY?
-------
PARPG- Indy post-apocalyptic roleplaying game currently in early planning stages.
It’s possible that a more complicated predictor that uses GB/FB/LD + BABIP rates exists, but I am not sure it has better results than FIP.
It’s called tRA, and you ca get it here.
Adoptive parent of Noah Lowry. Because he was awesome once, and, goddammit, he shall be awesome once again!
I hope.
FIP measures
what the pitcher controls. Or at least it attempts to isolate those things under the pitcher’s purview. It is “fielding-independent” pitching, which tries not to penalize a pitcher for having a weak fielding team. Having an Andruw Jones-caliber CF really helps reduce runs scored by turning fly balls into outs. A CF with less range would not get to those balls and they’d be hits, thus increasing the run-scoring chances for the batting team. No stat is ideal, but the goal of a stat like FIP is a worthy one.
Todos somos Gigantes
See here for a stat that incorporates some of what you’re looking for.
This seems to be more along the lines of what we should be looking for. Thanks for the link.
@Kenshin – Thanks, can you elaborate on what else is included in the formula?
@Zenbitz – That’s a good point about the difficulty of taking defense out of the equation. However, the end result of FIP is a pretty shitty stat if you’re looking for a backwards-looking explanatory statistic. As I said, I get the future predictor/forecasting aspect of FIP (I happen to know quite a bit about statistical forecasting), but the nature of forward-looking and backwards-looking statistics are quite different. Just because something is a better predictor of future success does not make it suitable for explaining the past (like win values is attempting to do).
@MarkOC – I have a hard time believing the ONLY thing a pitcher controls is K’s, BB’s, and HR’s. Another example where this comes up short – let’s say over a given time period a pitcher induces a higher % of IF flies than would normally be expected. These are automatic outs that should be solely credited to the pitcher. Just because a pitcher cannot necessarily keep this up in the future does not mean they should not receive credit for it over this time period, because after all, it did happen and it did contribute to winning games.
Basically some good feedback, my conclusion is FIP is a pretty shitty statistic FOR THIS PURPOSE, but thanks to Evan for giving us a better stat.
by Missing Barry on Apr 29, 2009 12:30 PM PDT up reply actions
Is there a good correlation from year-to-year for IFFB’s? I thought I read somewhere that it was really low but I can’t find anything on that right now.
My guess would be it regresses to some mean that’s a factor of FB% and K% with regards to AB’s. I haven’t read anything about it, though.
by Missing Barry on Apr 29, 2009 1:01 PM PDT up reply actions
does it matter?
I mean, sure it does if you’re trying to predict future performance, but whether a pitcher can keep up a ridiculous IFFB rate doesn’t affect how many he’s already gotten.
Sharlon Schoop - de favoriete Nederlandse honkbalspeler van McCovey Chronicles.
You always have to be one step ahead of your drunk friends
--Daisy Owl
Exactly. It matters for forward looking stats if your intention is to forecast a players production, but if your intention is to explain the past it doesn’t seem to me it should matter if it’s unsustainable, or even a skill, as long as the pitcher accomplished it.
Thanks for the follow-up Kenshin. My take – that’s pretty weak for the pitcher side of things. I hope they improve it in the future, although if these stats are being used for management decisions on players future performance, I can see the reasoning behind it…
Personally I think the most underrated stat is WPA. If I had a MVP/CY vote WPA would be the biggest factor I would use (though obviously not the only one). I think it’s a great idea for backwards looking explanatory purposes, though I haven’t looked through the details to see how well it allocates each play to the deserving player. I’d like to see it used more frequently.
by Missing Barry on Apr 29, 2009 2:25 PM PDT up reply actions
I still think WPA
rewards opportunity too much.
FIRE BRIAN SABEAN... UNLESS HE KEEPS DRAFTING WELL. .. AND SIGNS UNDERRATED PLAYERS LIKE AFFELDT OR PHELPS. .. OR ALRIGHT WHO'S PLAYING WITH THE ALIEN MIND-SWITCHING RAY?
-------
PARPG- Indy post-apocalyptic roleplaying game currently in early planning stages.
I have a preference for stats that are actually predictive. Isn’t that what they are ultimately for, anyway?
Randy Winn is going to catch that. And he'll do it real classy-like too.
not really
although you can have your preferences as you see fit. Let’s say you want to decide if someone deserves the MVP or to be elected in the HOF… you wouldn’t use predictive stats for that.
FIRE BRIAN SABEAN... UNLESS HE KEEPS DRAFTING WELL. .. AND SIGNS UNDERRATED PLAYERS LIKE AFFELDT OR PHELPS. .. OR ALRIGHT WHO'S PLAYING WITH THE ALIEN MIND-SWITCHING RAY?
-------
PARPG- Indy post-apocalyptic roleplaying game currently in early planning stages.
I have a preference for explanatory stats. Neither is necessarily right or wrong, you just have to know when the appropriate time to use them is. Predictive stats are forward looking, explanatory stats are backwards looking.
I see your point with WPA rewarding opportunity (they do have WPA/LI), but I kind of like that about the stat. It builds the whole “Player X played for a winning team” argument into the stat itself, or “Player Y single-handedly carried their team”. Of course, I don’t advocate solely looking at WPA – I think it’s a great stat and strong enough to be my primary stat when looking at something like an MVP race but definitely needs to be taken within the context of other stats.
by Missing Barry on Apr 29, 2009 4:09 PM PDT up reply actions
I guess
WPA is cute in a “replacement for pitcher wins, saves, and RBIs” sort of way
FIRE BRIAN SABEAN... UNLESS HE KEEPS DRAFTING WELL. .. AND SIGNS UNDERRATED PLAYERS LIKE AFFELDT OR PHELPS. .. OR ALRIGHT WHO'S PLAYING WITH THE ALIEN MIND-SWITCHING RAY?
-------
PARPG- Indy post-apocalyptic roleplaying game currently in early planning stages.
Well that just isn't a fair portrayal at all...
I’m too lazy to look too far into it, but I imagine there are some issues with who gets the credit for each change in winning probability, and certainly it’s missing the defense aspect (I’m assuming it only gives credit to pitching and hitting), but other than that it’s a fantastic idea that’s way more advanced and explanatory than things like wins, saves and RBI’s.
I’ll also give you the fact that opportunity is a significant factor in it, but I don’t see that as a negative, more a personal preference type thing. Sure two players might be equal players and Player X looks better than Player Y because they had more opportunities to help their team win, but then again that does mean Player X contributed more towards winning.
by Missing Barry on Apr 30, 2009 10:37 AM PDT up reply actions
No
Predictive stats help judge how much a player will contribute in the future. Explanatory stats tell you how much a player has already contributed. They answer completely different questions. You wouldn’t want to use explanatory stats to predict future performance, but you also don’t want to use predictive stats to explain past performance.
Sharlon Schoop - de favoriete Nederlandse honkbalspeler van McCovey Chronicles.
You always have to be one step ahead of your drunk friends
--Daisy Owl
Just to elaborate on this: predicting the future is a very difficult thing to do. When forecasting, often times the best simple stats work best (think FIP). This is because the more variables you add to a predictive stat, each one adds an additional layer of error to the statistic, so generally the best forecasts use a couple of the most predictive variables and leave all the less influential ones alone, because the additional predictive value is outweighed by the additional error. On the flip side, the past has already happened, and at this point statistics in baseball have evolved to where we can explain almost everything that happened in the past. So the good explanatory stats take account of all the variables we know because there is no error – they happened and we can measure the effect of each one.
by Missing Barry on Apr 30, 2009 5:42 PM PDT up reply actions
Better predictive stats have certain traits that make them better for “explaining” things in a statistical context (such as consistency and reliability). I can certainly see how there are situations where an explanatory stat with no predictive power would be useful, but other things equal I’d rather use a stat that can be used in a predictive model.
Randy Winn is going to catch that. And he'll do it real classy-like too.
The thing is though, it’s not like you have to pick one or the other. You can use both.
Sharlon Schoop - de favoriete Nederlandse honkbalspeler van McCovey Chronicles.
You always have to be one step ahead of your drunk friends
--Daisy Owl
Well it’s not like the “explanatory” stats have no predictive power – often times they actually have more predictive power, it just comes with a much larger error that makes it not worth using for forecasting purposes.
Basically think of a model where the variable in question is influenced by a bunch of other variables. Let’s use a pitchers performance as an example. It’s influenced by a number of pitcher controlled variables like K%, BB%, HR, GB%, FB%, etc. It’s also influenced by things out of a pitchers control – defense, park factor, BABIP, etc. FIP, as a predictive stat, takes only the most influential (and pitcher controlled) of these stats to use – K%, BB%, HR%. tRA expands on this, and does a better job of explaining a pitchers past performance since it has more pitcher controlled variables. When it comes to predicting the future, though, even though it has more variables to explain more, each variable adds additional error and at some point the added error outweighs the added information from the variable (that point is probably after K%, BB%, and HR)…
So the point is in this case FIP is better for predicting the future because it has almost as much predictive information with a lot less error than tRA, but looking at the past, tRA gives us more/better information. So like Viliphied said, use both!
by Missing Barry on May 2, 2009 7:55 AM PDT up reply actions
I was actually wrong...
the formula is FIP based: here is the explanation/justification:
Flossing a dead horse
Samples
wOBA may be a great stat, but there are still simple sample size funny business. Ishi’s “big” day yesterday of 2 hits, 1BB and 1 HBP plumped his RAA to -4.0. Russell Martin’s RAA is -3.9. We “know” Martin’s will go up. I know Ishi’s will.
Martin has had 80 something PA and Ishikawa has had 50 something. When each day brings another 3-10% bump in PA, the sample will change. Travis “halved” his Win value from -.4 to -.2.
co-dad of IshikaBOOM w/AfDC.
Ishikawa, let the boy hit against lefties.
It's amazing that...
Sabes and Boch’ do whatever they can to force feed Eugenio Velez when he does little good on the baseball field. Sure he can run, but that’s not valuable when you don’t get on base. It’s also poor roster management to carry a pinch running specialist when there’s a dearth of offensive production on the team. It’s also poor managing to play a dreadful defensive player at premium defensive positions.
So after our long discussion about FIP, predictive stats, explanatory stats and such, I find that I read this particular article on fangraph, but apparently forgot to read the important parts related to our discussion. To quote the article (explaining why FIP is used in fangraph pitcher win values):
“That doesn’t mean that we’re set in our ways and that these win values will never be improved upon. If and when a new metric like tRA is proven to be significantly more effective in valuing pitchers (and I’m hopeful that it will be, given more data exploration on the topic), we won’t be standing here as guardians of the infallibility of FIP. We want to get to the truth, and do so as quickly and as accurately as possible. I will encourage you (especially those of you in the "tRA is awesome/FIP sucks" camp), though, to not let minor differences cause you to miss the fact that FIP and tRA lead to very similar results. "
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
Certain photos copyright © 2012 by Associated Press or Getty Images. Any commercial use or distribution without the express written consent of Associated Press and Getty Images is strictly prohibited.
