Friday, February 15, 2013
This post on Wages of Wins says that Plus/Minus is bad. And I'm telling you it's a great stat, if adjusted. That post says:
Indeed, we’ve seen iteration after iteration of “Adjusted Plus Minus” to improve upon the fact that all prior versions are bad at explaining wins. It’s a self-fulfilling prophecy. If you bring complicated metrics to try and solve a problem, you will remain convinced the problem is complicated and thus be convinced the only thing to solve the problem with is a complicated metric.
While I like plus/minus in hockey, I love it in basketball. There's not enough scoring in hockey to get good samples, which is why we look for anything, like shots for and against. In basketball, there's scoring a-plenty.
Anyway, this is the actual fact: we are recording who was on the ice/court/field when a run/goal/point is scored. We are also recording this when you are NOT on the field of play. This is actually at the heart of WOWY (with or without you). And, as we've seen, it's fantastic. And others have used that same process for cathcer framing. We use it all the time with ERA: we compare the pitcher's ERA to the rest of his team or rest of the league. We do it with park factors. Patriot has also done it with an adjusted "Wins Above Team". It works fantastically well.
So, the idea of plus/minus is as old as analysis itself: what happens when the player is involved and what happens when he is NOT involved.
Now, the issue occurs when going from "involved" to "responsible". You could in fact look at TEAM BABIP for a team when Andrelton Simmons is, and is not, playing SS. This is a form of plus/minus. The issue is of course that Simmons may be on the field when alot of balls are hit, except he's not responsible for most of them. We solve this problem by putting Simmons on the field for 50,000 balls in play. Unfortunately, that's a long career's worth, and we can't wait that long.
So, that's what we have: we have a signal, and we have the noise. We make our adjustments to reduce the noise so that that signal can come through.
In 1986, Mark Howe was an astounding +85 (as was Brad McCrimmon equally high). The Flyers were +74, meaning that when Mark Howe was not on the ice, the rest of the team was a negative. Now, we don't know if it was Howe and/or McCrimmon, or the rest of the defensemen were terrible, but what we do know was that it was an enormous gulf. There's signal in there. There's noise too.
Which is why we live for sample size. Which is why the work I did with catchers is as powerful as it is. Not only do I have the sample size, but I also have the benefit that catchers have been exposed to dozens and in some case over a hundred, different pitchers in their careers.
If you don't like plus/minus, fine, whatever. But calling it bad means you've come to a conclusion. And that conclusion won't hold.
There are undoubtedly two marquee events at the Winter Olympics: figure skating, and men's hockey. For figure skaters, the Olympics is a tremendous marketing opportunity for them. They can parlay that exposure into skating "tours". Hockey however is much different. They already have their own league. The exposure helps when they can show the games live and in normal viewing times. Socchi, Russia doesn't help them so much.
In addition, the IOC has broadcast and distribution rights. Whatever "Olympic ideals" is supposed to exist in the utopian world has to come face-to-face with the "absolute power corrupts absolutely" with our real-world Bayesian prior. The NHL, and NHLPA, want their cut of the action. So, now it seems they're trying to figure out the value for each party:
In return for sending its players to the Sochi Olympics, the NHL is trying to acquire video, photograph and website rights for the games. The IIHF and the IOC retain those exclusive rights now.
The NHL began sending its players to the 1998 Games in Nagano, Japan, and continued through the 2010 Vancouver Games. Even though the NHL received great exposure by having its players take part in an Olympics in North America, disrupting the season does come with a cost.
The stopping of the season, the potential injury risk to players, and no tangible upside for the NHL are all factors that create doubt about whether the investment is good for the league.
?
(11)
Comments
• 2013/03/03
•
Hockey
Thursday, February 14, 2013
Using my Markov, we end up with these values. x-axis is the OBP, and the y-axis is the run value relative to all PA.
Wednesday, February 13, 2013
First GM from Europe.
Kekalainen, who spent eight seasons with the NHL's St. Louis Blues from 2002-10 where he was involved in all facets of hockey operations, takes over a Columbus team that has not made the playoffs since 2009 and finished last overall in 2012.
Prior to working for the Blues, Kekalainen spent eight years with the Ottawa Senators as their top European scout and later as director of player personnel. He also played three seasons in the NHL with the Boston Bruins and Senators.
His resume reads like a standard resume of a GM, except for his birthplace?. Given the huge number of European players in the NHL, it's kind of weird that he's the first. I don't know how many hundreds of people have been GMs in the history of the NHL, but the odds that all of the best qualified had to be Canadian or American seems to be a stretch.
Similarly, how long before a woman takes the leap too? They're already part of the CHL. Hockey seems to be different from the other sports, as being a former player seems to be a pre-requisite. Which tells me that it's either a highly specialized job, or it's likely a very inefficient selection process.
Jon continues his work on calibration and adjustment.
A contract extension is when a player's current deal is extended in terms of years, and usually with a higher rate per years. Felix went from a 2/40 deal to a 7/175 deal.
But, let me make a crazy suggestion: what if it went the other way? Not an extension, but a ... flexion. Verlander also has a 2/40 deal. What would it take for both sides to agree to rip up that deal, and sign a one year deal (MLBPA notwithstanding). That is, if they agreed to a 1/10 deal, and then Verlander is a free agent, would both sides go for it? How about a 1/5 deal? 1/1?
At what point does it make sense??
(20)
Comments
• 2013/02/17
•
Finances
Steve does a good job by going through some various issues when using Linear Weights.?
Carson gives us an update. ?
It's blocked at the office, but if someone can post a few interesting tidbits, that'd be cool.?
Monday, February 11, 2013
Discuss this in our forums
Russell talks chemistry.
It should be clear that since we are dealing with players, and players are humans, then chemistry exists. This however is not the issue. The issue is the extent to which we can identify it, and the extent to which we can measure it. This is what saberists do. Identify and measure. This is our job. Clutch exists because humans are involved. Chemistry exists because humans are involved.
If we can't identify it, how can we measure it? So, that's why saberists don't include it as a parameter: they are not able to identify it.
Teams however still try. I noted four years ago when Cliff Floyd was given a contract, even though he was a replacement-level player.
When people say that you can’t measure the intangibles, remind them that those intangibles are being paid for with tangible dollars. And the value of those intangibles, as determined by MLB, is $350,000. If it was worth more, then some team would have paid more. They didn’t.
So, the point is how much do you pay for that chemistry that you have a hard time identifying and hard time measuring?? If one player loaded with chemistry and intangibles is going to generate 20 wins over the next five years by himself, and X wins that the rest of the team is going to gain because of his synergestic chemistry, then do you pay him more or less than a player that generates 25 wins over those same five years, but has intangibles that are divisive and contributes to unhappy teammates?
Yes, you should identify it, and yes you should measure it. But, how far are you going to go? How much are you going to pay to get a worse player that has great chemistry? And what if that divisive player ends up being Reggie Jackson, Albert Belle, or Barry Bonds?
Everything has its price.
(25)
Comments
• 2013/02/11
•
History
Sunday, February 10, 2013
?I don't know if/when they will be released.
I published the formula on March 10, 2004, and it has not changed at all since then. I'd say anyone who wants to create a forecasting system should first try to replicate that.
(7)
Comments
• 2013/02/12
•
Forecasting
?Several months back, I was arguing that the NHL needed to implement a Cap Benefit Recapture into the CBA, that it was incredibly silly they way they used to have it. Well, now they did it, but it seems there's some huge misunderstanding.
The way it should have always worked is this: if a player retires early, and ended up earning say 100MM$ in salary, but 90MM$ in cap-hit, then there's 10MM$ of cap-hit that was deferred. Since he retired early, it's time to pay the tax man. If he retired 2 years early, then you spread that 10MM$ deferral over those two years, and he gets a 5MM$ cap hit each year.
Simple and natural, something any accountant would do. So, what's the problem?
The problem arises with traded players. If a player is traded early in his contract, such that a team might have paid a player 70MM$ in contract, but 40MM$ in cap-hit, there's 30MM$ of deferral. But, the player continues to play for his other team. That team will eventually have years were the cap-hit is higher than the contract paid. THAT EXTRA HIT REDUCED THE DEFERRAL OF THE ORIGINAL TEAM. I do not know this for a fact, but it seems plainly obvious that this is what has to happen.
The entire point is that once a player retires, you look at his total contracts paid, and his total cap-hits charged. The difference is the deferral that has to get paid. And that gets charged to the team that traded him. So, it's ridiculous to think that if Weber retires one year early that he'll have a 32MM$ cap charge. That's because then he'd have a greater cap charge than salary.
So, it's very simple: lifetime contract = lifetime cap-hit. That's it. That's the only thing you have to remember. If there's a shortfall, it's THAT shortfall that gets charged. It's not the interim shortfall when he was traded. That may be the STARTING POINT, but that amount eventually gets reduced as the years go on.
(3)
Comments
• 2013/02/12
•
Hockey
Saturday, February 09, 2013
Good job by Franklin.
Specifically in Bob’s case he’s confused the fact that more ‘flashy’ drivers tend to buy flashy red sportscars, and are more likely to drive ‘flashy’ and pass somebody in the suicide lane during rush hour to skirt through a changing light, and get into a wreck. The red paint and more accidents are both effects of the same root cause (wanting to be flashy while on the road), but Bob has assumed that the paint’s ability to absorb high frequency light waves is causing more accidents somehow.
...
Thus, it’s often the case that winning is causing rushing yards, and losing is causing a lack of rushing yards, instead of the other way around. The team is running because they’re winning, not winning
because they’re running.
?Adam gives his treatment to Rick Reuschel.
As of right now, Reuschel, according to the readers of Baseball Reference, is 84th best pitcher, between David Cone and Chuck Finley. Smells about right. He's also ahead of Fergie Jenkins and Jack Morris and El Presidente.
(17)
Comments
• 2013/09/16
•
History
?For those wishing to start their own threads, I have a discussion board setup.
http://tangotiger.com/index.php/boards/
?Someone is taking the initiative.
(8)
Comments
• 2013/04/10
•
Poll
Recent comments
Older comments
Page 148 of 151 pages ‹ First < 146 147 148 149 150 > Last ›Complete Archive – By Category
Complete Archive – By Date
FORUM TOPICS
Jul 12 15:22 MarcelsApr 16 14:31 Pitch Count Estimators
Mar 12 16:30 Appendix to THE BOOK - THE GORY DETAILS
Jan 29 09:41 NFL Overtime Idea
Jan 22 14:48 Weighting Years for NFL Player Projections
Jan 21 09:18 positional runs in pythagenpat
Oct 20 15:57 DRS: FG vs. BB-Ref
Apr 12 09:43 What if baseball was like survivor? You are eliminated ...
Nov 24 09:57 Win Attribution to offense, pitching, and fielding at the game level (prototype method)
Jul 13 10:20 How to watch great past games without spoilers