The problem there is that the implication would be that during the learning curve, they would persist in giving away or not persuing superstars. If true that's disastrous and has long term consequences even if you do one day wise up. Its just too hard to get them to blow even one opportunity.
But again, not one single thing this front office has said cleanly jives with their actions. Want to improve defense. Instead get rid of players with defensive potential and add ones without. Want to increase passing, add Carl Landry. Tag SF as a priority. Ignore SF. Say we need to keep salary cap flexibility, then sign Carl Landry to a stilly 4yr deal, try to dump giant money on Andre Igoudala, and now maybe the same on Monta Ellis. All fo those latter moves directly conflicting with "this year won't be about wins and losses" as well. So if now they imply they want to overpay bench players, only to then start throwing big money at Monta Ellis...well, its hardly surprises me.
Do I need PowerPoint or Excel to claim I'm building a successful team based off of analytics? I can never remember. What is the metric used to verify the validity that one is a 'numbers' guy?
Is it looking at charts? Filtering columns from highest to lowest? Or do you have to be able to enter a function into a cell?
And once I've established that I am in fact a 'numbers' guy - how many additional wins do I need to get before it has any meaning whatsoever? 5 games?
I'm getting a little excited about my potential career change. Or, at the very least, furthering my current career by yelling 'I'm moneyballin' it bitches!’ after every time I do something stupid.
Do I need PowerPoint or Excel to claim I'm building a successful team based off of analytics? I can never remember. What is the metric used to verify the validity that one is a 'numbers' guy?
Is it looking at charts? Filtering columns from highest to lowest? Or do you have to be able to enter a function into a cell?
And once I've established that I am in fact a 'numbers' guy - how many additional wins do I need to get before it has any meaning whatsoever? 5 games?
I'm getting a little excited about my potential career change. Or, at the very least, furthering my current career by yelling 'I'm moneyballin' it bitches!’ after every time I do something stupid.
I seem to remember the moneyball nerd was preoccupied with "on base percentage" no matter where it came from.
Maybe our group is trying a similar idea with "assists". They figure if you can slap a group together that can put up 30 assists per night they will always win (Koz's rule).
It's working in American football and in basketball, don't know about hockey. The teams who are winning in the NBA use moneyball style thinking right now. Same with the NFL (see Patriots and 49ers as examples).
It's obviously easier in baseball, with more games and more discrete events. Top hitters get 500+ at bats every season, teams get 162 games, etc.
In basketball there are 82 games, which makes it easier than in american football, but because the play is continuous and the play of all five guys on each team affects the outcome of the possession nearly every time, it is harder to get more detail. But that doesn't mean it doesn't provide useful information. Teams are getting plenty of data and using it to improve.
Moneyball is about finding where the value is, finding out what is overrated and what is underrated in the league and exploiting those pieces of information. Daryl Morey is absolutely playing Moneyball. He is the poster boy for analytics in the NBA and he hasn't won anything yet, but his Moneyball thinking led him to understand that there's value in amassing assets to try to get superstars (the people who misunderstand Moneyball are the ones who think he's not doing Moneyball). We'll see how Howard and Harden do, but the salary cap and maximum contracts make superstars worth more than their contracts. Exploiting that is exactly what Moneyball is about.
Similarly, the Patriots have been using Moneyball style thinking for years, and it has kept them on top. Over the last few years, the 49ers have done the same (one of their top guys is a regular Sloan Sports and Analytics Conference attendee/presenter) and because of that they are already at the top of the NFL and poised to stay there.
Some people think Moneyball or analytics is about only looking at numbers, but of course that's not what Beane did (nor what he continues to do) and it is not what current analytics guys do and hopefully it's not what Ranadive wants to do.
Theo Epstein says, "huh?" Brian Sabean says, "wha?"
I keep hearing the Athletics being cited as an example of Moneyball "working." I'm a casual baseball fan, at best (which is to say, when the Cubs aren't any good, I could give a **** about baseball), so maybe someone could help me out here: how many World Series have the Athletics won in the Beane era? When's the last time they even got to the ALCS? I mean, I realize that, in Major League Baseball, just getting to the playoffs is a much bigger deal than it is in the NBA, but exactly what definition of "success" are you guys working with? If you have a strategy that gets only gets you to the front yard, and then another guy comes along, with a bigger checkbook, and uses a tweaked version of what you did, with better players, and wins the whole thing, does that really prove that what you did works?
Reading over this again, I think the only point we really disagree on is the definition of what moneyball is. My point was more that moneyball was primarily concerned with finding ways to win in a small market and one way that was done was with advanced statistical analysis. And that's become a kindof shorthand that people use, referring to statistical analysis as moneyball. But the thesis of the book was that a team can still win without spending money by identifying and exploiting market inefficiencies and that hasn't really proven true. The big spenders are still the ones winning. Not all the time, it's still entirely possible to blow a lot of money without getting the expected wins, but at the playoff level when you're down to the elite teams, the benefits gained by maximizing your resources isn't enough to overcome the spending gap. Everything has to break right for the A's to win. Other teams have the luxury of winning despite their redundant pieces and overpaid poor performers.
That said, statistical analysis has been a huge success in baseball and that's why it's filtering out now into the other sports leagues. And it is common sense that utilizing more tools to make your decisions can give you an advantage, provided you know how to use them.
Agreed. I tend to think of Moneyball as usable in big markets as well, since you can find market inefficiencies in Boston and New York and you still have constraints on things like how many players you have on your roster, so the principles can still apply.
I keep hearing the Athletics being cited as an example of Moneyball "working." I'm a casual baseball fan, at best (which is to say, when the Cubs aren't any good, I could give a **** about baseball), so maybe someone could help me out here: how many World Series have the Athletics won in the Beane era? When's the last time they even got to the ALCS? I mean, I realize that, in Major League Baseball, just getting to the playoffs is a much bigger deal than it is in the NBA, but exactly what definition of "success" are you guys working with? If you have a strategy that gets only gets you to the front yard, and then another guy comes along, with a bigger checkbook, and uses a tweaked version of what you did, with better players, and wins the whole thing, does that really prove that what you did works?
If you want a true test, you have to determine the criteria for success before you start. We can't go back and do that, so we have to try to assess its impact based on what we know happened, without being able to test any counterfactuals.
When Beane (and Alderson to some extent before him) started using the "Moneyball" techniques, the A's had one of the lowest payrolls in the game. To many, simply competing for a championship would be considered a success for that team. That is what they did in the era in which they implemented "Moneyball". I'd consider that a success.
If you prefer not to consider that a success, and demand rings, then of course you should look to the Red Sox, who had a huge payroll but kept failing to win. They then turn to a tweaked version of the A's "Moneyball" ideas and promptly won two World Series.
Now, if you want to say that somebody else tweaking the "Moneyball" idea and actually achieving championships with it shouldn't count as success, and it's only a success if the original "Moneyball" team had won a championship, then fine, it wasn't a success. Seems like a a stringent test for success, and not terribly relevant to forward looking questions, but if that's your criteria, then it fails.
To me, the best measure of success is whether you achieve better results with your strategy than you would have otherwise given the same circumstances. I think the A's have done better than most small-market teams since they started using "Moneyball". I also think that not long after they did that, other teams began using similar techniques, so the success they continue to have with a small payroll isn't so much an indication of the success of "Moneyball" as it is an indication of the quality of the A's organization in general (front office, scouting, etc). Now so many teams use sabermetrics and data to attempt to find hidden value that it's difficult to differentiate which ones do what. But as I said earlier, one could certainly claim that that in and of itself satisfies a criteria for success.
When Beane (and Alderson to some extent before him) started using the "Moneyball" techniques, the A's had one of the lowest payrolls in the game. To many, simply competing for a championship would be considered a success for that team. That is what they did in the era in which they implemented "Moneyball". I'd consider that a success.
If you prefer not to consider that a success, and demand rings, then of course you should look to the Red Sox, who had a huge payroll but kept failing to win.
Now, if you want to say that somebody else tweaking the "Moneyball" idea and actually achieving championships with it shouldn't count as success, and it's only a success if the original "Moneyball" team had won a championship, then fine, it wasn't a success. Seems like a a stringent test for success, and not terribly relevant to forward looking questions, but if that's your criteria, then it fails.
To me, the best measure of success is whether you achieve better results with your strategy than you would have otherwise given the same circumstances.
For baseball (at that time especially), it's pretty much getting to the playoffs or being on the cusp of the playoffs. I'd say it would have to be done consistently over a period of years to be considered evidence of success. There's also an element of "legitimate chance to win", but in baseball that's often anybody who makes the playoffs.
My standard of success is not better than yours. And why should anybody have to decide? We're not trying to make an objective determination of the one true answer to the question, 'Was Moneyball successful?'
What we're trying to do is assess Moneyball to see whether we as individuals consider it successful in order to place it in historical context, determine whether it was a strategy worthy of emulation, or simply as an interesting exercise in internet discussion. It doesn't matter whether the final determination of success comes out "yea" or "nay", that's irrelevant. What's relevant is why.
So, again, the answer to the question depends on what you mean by "success". Were the Webber Kings successful? Depends on what you mean by success. You tell me your definition of success and I'll tell you whether I think they were or not. By my definition of success in most contexts, I'd say they were, in that they contended for a championship and had a legitimate chance at winning one. To me, the efforts made by the front office to build that team were successful in that they were worth emulating and should be looked upon favorably. I think the same thing applies to the A's, and in the context of "Moneyball" I think their results indicate that their strategy was worth emulating and should be looked upon favorably from a historical perspective.
A common theme in the stats/data world is that it's the process that counts more than an individual outcome. If you do what appears to be the smart thing and it doesn't work out, then don't assume you did the wrong thing based on the one attempt. Trading a solid star in Richmond for a potential superstar with perceived flaws in Webber was the right decision in my opinion, but if your criteria for success is rings or conference championship banners then it failed. Does that mean you shouldn't try it again? Of course not. It got the team closer to winning than it had been in decades. A bounces a different way one time and we all would have been celebrating. The same is true of the A's of that era. A ball bounces a different way one time, and they could easily have won it all. Is that success?
My point earlier was that if you look at the A's payroll you would expect them to finish in last place in their division every year and that's not happening. They're winning a lot more games than they're losing and they're doing it without spending money so something is working there. But I was also clear to point out that most people measure success in any league not just by games won but by championships -- how well do you perform in the high pressure situations against elite teams? And by that measure the A's just aren't on that level. So that's why I said it's debatable that Moneyball has been a success. The parts of it that are working -- besting your competition on player evaluation and scouting -- have been adopted by the rest of the league. And when it comes to re-signing your own stars or filling holes in the free agent market, the A's are so far off the mark because of their market limitations that it's really the best they can hope for just to make the playoffs and hope to get a little bit lucky. Now if you have both competitive advantage in payroll budget and competitive advantage in scouting and analytics, than you are a real threat to win it all. But how much of that is moneyball and how much of it is good old fashioned "this guy hits a lot of homers, let's give him $20M a year"? The point is that finding those under valued performers can improve your team, but baseball too is still a league dominated by stars. This season the A's have found themselves some cheap ones in Josh Donalson, Bartolo Colon, and Grant Balfour. If they win it this season Beane will get his due (here's hoping!) but it hasn't happened yet.
But I don't think uolj was saying that the A's are a success before. He was saying that the spread of statistical analysis to every front office in baseball is an indication that the information it provides has value. Moneyball itself may not have been a success (yet ) but the ripple effect that it has had throughout the game has to be seen as evidence of a tool that is working for what it's supposed to do. OPS (On-Base plus Slugging) is now considered the stat for measuring how successful a hitter is over a season. And there are all sorts of "ballpark independent" stats which can be used to evaluate pitchers on a more even playing field. Finding equivalent types of tools in other sports is a lot more difficult than it sounds on the surface though. Baseball is just a really unique type of game that lends itself to this kind of analysis in a way that two teams playing offense and defense simultaneously on a 92 foot long court for 48 minutes does not. Different types of stats like +/- and PER are attempting to provide the same types of objective analysis but I don't think there's really a consensus yet about what works and what doesn't.
as far as NBA things are concerned, does anybody have an idea who of the GMs in control of a team since this advanced stats movement has been around was actually an advanced stats guy? I know Morey is and R.C. Buford is as well (had his own panel at Sloan Conference and everything). apart from those two? Sam Hinkie clearly qualifies, too, but I have no idea if, say, Masai Ujiri, Kevin Pritchard, Bob Myers or Sam Presti are part of the group, too.
