 |
APBRmetrics
The statistical revolution will not be televised.
|
| View previous topic :: View next topic |
| Author |
Message |
John Hollinger
Joined: 14 Feb 2005
Posts: 175
|
 Posted: Mon Dec 17, 2007 10:21 am Post subject: |
 |
|
Indeed, I don't have a Ph.D. in Economics -- that was my road not taken. I was about thisclose to going to Washington University in St. Louis to do just that, but decided to move to Portland and listen to some grunge instead. I do have an undergrad degree in Econ, and I think that an economist's mindset is a very important tool in this field, even if it comes without the classwork.
As for where we stand in this field, I think we're absolutely at the tip of the iceberg. Whether it's five years from now or 50, I fully expect that somebody will come up with a superior methodology to PER, for instance, and I fully expect that the data available to us will grow by leaps and bounds, especially now that the league finally seems to be taking an interest in stats.
As far as the Berri thing, I'm of a mixed mind. I do worry that people, especially those who are hostile to stats to being with, will look at it and say, "you see, this stats stuff is all a bunch of crap", which decreases the chance of more valid work gaining acceptance.
On the other hand, I do worry that mentioning it so much only provides more free publicity and calls more attention to what is, in the opinion of myself and many others, a very deeply flawed approach.
Basically, I'm still not sure what the right answer is. Which I guess kind of explains why I've split it down the middle so far. |
|
| Back to top |
|
 |
asimpkins
Joined: 30 Apr 2006
Posts: 245
Location: Pleasanton, CA
|
| Posted: Mon Dec 17, 2007 12:42 pm Post subject: |
|
|
I agree that there's a lot of important work outside of player valuation, and I agree that the biggest gains will come with counting new things, but I'm not sure why it has to happen at the exclusion of player valuation discussion.
Some people get a lot out of it for a variety of reasons and some people apparently don't. Couldn't those that don't just avoid the threads they aren't interested in and start threads that they are? |
|
| Back to top |
|
|
HoopStudies
Joined: 30 Dec 2004
Posts: 706
Location: Near Philadelphia, PA
|
| Posted: Mon Dec 17, 2007 1:22 pm Post subject: |
|
|
| asimpkins wrote: | I agree that there's a lot of important work outside of player valuation, and I agree that the biggest gains will come with counting new things, but I'm not sure why it has to happen at the exclusion of player valuation discussion.
Some people get a lot out of it for a variety of reasons and some people apparently don't. Couldn't those that don't just avoid the threads they aren't interested in and start threads that they are? |
Definitely and that's generally what I personally mostly do. I highly recommend it as blood pressure therapy in fact.
I would like to think that we weren't spending 90% of our effort (which is more representative of the number of posts on player valuation) on 10% of the game. But that is very clearly the interest balance here.
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers. |
|
| Back to top |
|
|
NickS
Joined: 30 Dec 2004
Posts: 384
|
| Posted: Mon Dec 17, 2007 1:36 pm Post subject: |
|
|
| Quote: | | I would like to think that we weren't spending 90% of our effort (which is more representative of the number of posts on player valuation) on 10% of the game. But that is very clearly the interest balance here. |
Look, there's an obvious reason why Player Evaluation (and methods of evaluation) are a mainstay of discussion around here: it's fun, easy, interesting, and an easy way to relate the abstrations to real basketball.
It's easy because, ideally, the results of a player evaluation post don't require any specific knowledge to understand (though it takes more to understand the details of how those results were produced) so it's something that everyone can kick around and kibbitz.
It's interesting because it's one way to make sense of all of the new data that's coming in during the basketball season -- who's playing well, who's playing poorly, which teams are doing well, which teams are outperforming their +/-.
For the same reason it's interesting, it's a useful feedback loop between in-season results and APBRMetrics work. Can APBRMetrics say anything interesting about why things are turning out they way they are during the season?
So I don't think it makes sense to complain that there are too many posts and comments about player evaluation, the real question is are they crowding out other research, and that's a much more difficult claim to make. It may be, and that's dissapointing if true, but it's equally possible that all the discussion of player ratings is, on some level, background chatter that's always going to go on. |
|
| Back to top |
|
|
HoopStudies
Joined: 30 Dec 2004
Posts: 706
Location: Near Philadelphia, PA
|
| Posted: Mon Dec 17, 2007 2:08 pm Post subject: |
|
|
| NickS wrote: |
So I don't think it makes sense to complain that there are too many posts and comments about player evaluation, the real question is are they crowding out other research, and that's a much more difficult claim to make. It may be, and that's dissapointing if true, but it's equally possible that all the discussion of player ratings is, on some level, background chatter that's always going to go on. |
Very reasonable.
I think when I looked at those numbers, I realized that it was like asking us to stop talking about sports. Ain't gonna happen.
_________________
Dean Oliver
Author, Basketball on Paper
The postings are my own & don't necess represent positions, strategies or opinions of employers. |
|
| Back to top |
|
|
Dan Rosenbaum
Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
|
| Posted: Mon Dec 17, 2007 3:03 pm Post subject: |
|
|
| No one has potentially benefited more from WoW than Denver. Wins Produced places very little value on a player Denver traded for (Allen Iverson) and a great deal on a player they traded away (Reggie Evans). These two players, in fact, are mentioned in the first two comparisons in the link that JohnH provided. Keep that in mind when you read these posts. Folks are always looking to create an advantage any way they can in a highly competitive business like the NBA. |
|
| Back to top |
|
|
Flint
Joined: 25 Mar 2007
Posts: 112
|
| Posted: Mon Dec 17, 2007 5:03 pm Post subject: |
|
|
Dan - How do you mean Denver has benefitted from the WOW? Did they benefit by getting Steven Hunter, by unloading Evans' contract, or by getting Iverson for Miller?
If the latter, I am sort of suprised you would say that, since your adjusted +/- results suggest Miller has been a better player, as the WOW indicates as well.
Denver is having an interesting season, that is for sure, 13th best offense, 2nd best defense. |
|
| Back to top |
|
|
mtamada
Joined: 28 Jan 2005
Posts: 377
|
| Posted: Mon Dec 17, 2007 6:18 pm Post subject: |
|
|
| Harold Almonte wrote: | | About the baseball comparison: Basketball could be most predictable than baseball even with worse data and metrics, I think it's because the player's slumps are not as long, and un-make up-ble at the replacement, I think probably the needs of testhosterone are different in both games, but that's another threat. |
The predictability issue is an interesting one, that was one of the points that Bill James made in his critique of the NBA a couple of months ago: that the results of NBA games (and seasons) are too predictable, resulting in reduced fan interest.
I don't know if he's correct or not, but one of his points that makes sense is that there are a lot more events in a basketball game than in a baseball game (well, maybe about the same if we count every single pitch in a baseball game as an event). I.e. in evaluating a batter, even after an entire season we've typically got fewer than 600 plate appearances to look at. And maybe only 5 in one single game.
In the NBA, a high usage player might have 1,500 field goal attempts in a single season -- and 20 or more in a single game.
So we have larger sample sizes in the NBA for many of the analyses that we do. And that may, to a small extent, counteract the well-known complexity that NBA stats have compared to baseball, where teammate interactions are much less important, and plate appearances are determined by the lineup, not by the players' decisions.
I don't know if that makes the NBA more predictable in the end, but it's a factor leading to greater predictability. |
|
| Back to top |
|
|
mtamada
Joined: 28 Jan 2005
Posts: 377
|
| Posted: Mon Dec 17, 2007 6:28 pm Post subject: |
|
|
| Flint wrote: | Dan - How do you mean Denver has benefitted from the WOW? Did they benefit by getting Steven Hunter, by unloading Evans' contract, or by getting Iverson for Miller?
|
My interpretation of DanR's quote is that he meant that Denver benefited by doing the opposite of what WoW would have recommended. I.e. to the extent that WoW has garnered interest, and to the extent that a team such as Philadelphia decides to follow WoW's guidelines, a team such as Denver will find it easier to pull off trades that are lopsidedly in their favor.
I presume that the question of how much Philadelphia has used WoW, if they've even read it at all, is hypothetical and speculative. I.e. I don't think we know that Philly literally did use it in their trade decisions. |
|
| Back to top |
|
|
gabefarkas
Joined: 31 Dec 2004
Posts: 1313
Location: Durham, NC
|
| Posted: Mon Dec 17, 2007 6:40 pm Post subject: Re: Next in Basketball Analysis |
|
|
| HoopStudies wrote: | | Football quantitative analysis is operating in a very different way than we in basketball or those in baseball operated - but a way we might consider. Aaron Schatz and his collaborators really do a great job at studying small things. They study the offensive line, the defensive backfield, the wide receivers -- the units. They better understand how those units operate without yet putting together the big picture of how the units interact. If you haven't seen Pro Football Prospectus, I recommend it as a different perspective on how to study sports at the very least. For us, we don't need a more detailed framework to understand parts of basketball that we aren't capturing right. When we say a person is a good rebounder or good passer or good shooter, what does that mean? We don't even have a good "passer rating." Roland's passer rating was quick and dirty and, I think, he doesn't really try to stand behind it. By coming up with a new passer rating, even conceptually without the numbers, it can lead to insights about a part of the game. |
I'm a big fan of Football Outsiders too, and I think some of their stuff is just great. However, in my opinion, football is a very different animal than basketball. Every single one of the 11 players on the field has a distinctly different role in football. Even within a position that role varies depending on the play, but the gist of it is basically the same, and to a large part, non-interchangeable.
Basketball is different. It's more fluid. Roles change, team-to-team, squad-to-squad, and play-to-play. Players' skills change over time.
Referring specifically to passer ratings, I think the definition of "good" here changes from team to team. As you adroitly mentioned, we'd be smart to stay away from subjectivity like "what is a 'good' passer". Also, like the contested shot and pick in basketball, or the block and quality of play action in football, making a good pass isn't recorded. Sure, we have the "assist". But to me, the notion of an assist, as it currently exists, is laughable. It's basically a subjective measure that somehow became an accepted part of the basketball reference terminology. You mentioned the need to collect more data; I think this is the first step: defining a standard rubric of what to collect and how to collect it. |
|
| Back to top |
|
|
Flint
Joined: 25 Mar 2007
Posts: 112
|
| Posted: Mon Dec 17, 2007 6:44 pm Post subject: |
|
|
| That is how I understood what Dan said also, which seemed curious to me given that his metric offered the same conclusion as the WOW in this case, that Miller is a better, cheaper player than Iverson. But perhaps there is some kind of synergy issue I am missing behind the numbers... |
|
| Back to top |
|
|
gabefarkas
Joined: 31 Dec 2004
Posts: 1313
Location: Durham, NC
|
| Posted: Mon Dec 17, 2007 6:51 pm Post subject: |
|
|
| Dan Rosenbaum wrote: | For the longest time you made the argument that player evaluation metrics could not be evaluated. But my paper with Dave Lewin does just that in two different ways. That to me is a big contribution that helps advance the field. Being able to evaluate what we do is really important and these methods we developed could be adapted to evaluate lots of other things.
Between our JQAS paper and this paper, we also make advances in the area of the theory of possession usage. Possessions are a fundamental building block of practically everything that we do, so I think this is useful. |
You keep mentioning this paper with Lewin, but either I missed something or it just kind of started coming out of nowhere. Is there an introductory post about it that I missed? Did you share a draft or something? Can you talk about it at all? Is there a draft abstract? Or is this the Pot/Kettle paper that is printed and in my bag and that I've been meaning to read for about a week now? |
|
| Back to top |
|
|
Mountain
Joined: 13 Mar 2007
Posts: 1527
|
| Posted: Mon Dec 17, 2007 7:01 pm Post subject: |
|
|
"Dan Rosenbaum wrote:
Finally, I think this whole discussion has really put a bulls-eye on the usage/efficiency tradeoff, and I think that is a good thing. Understanding that tradeoff better is a key to doing basketball statistics analysis better. "
So taking mental cue from that and looking at statistical +/- for other reasons I took notice of this passage in Dan R's original adjusted +/- presentation:
"At the mean level of effective field goal attempts, the marginal cost of another two point field goal attempt is equal to 1.09 points per 100 possessions, but at 25 effective field goal attempts per 40 minutes, the marginal cost is just 0.58 points per 100 possessions.� This declining marginal cost likely reflects the value of players who can generate field goal attempts as the shot clock is expiring or under extreme defensive pressure.� Players who attempt lots of three points and free throws appear to be more valuable than players who specialize in two point field goal attempts�"
I wonder if it would be possible to separate out crunch and clutch time shooting from high usage in the rest of the game? Is this the main or sole basis for the declining marginal cost or is there a usage / team efficiency relationship in these normal parts of the game apart from that? Equal or lesser? That would matter a lot in evaluating player mix and potential adds / subtractions. How many good crunch / clutch time capable players do you want, how will they interact, how much will they be used and how much impact will they have beyond the direct share of the load, the indirect rasing of team efficency?
Without having gone back to all previous discussion, I'll ask how should further research on the usage/efficiency tradeoff be framed? Usage vs eFG%, TS% or offensive rating? The later best? I think the study should look at breakdowns by not only what is prominent in shot location / possession usage charts but perhaps also size and / or position, experience and see what is found. |
|
| Back to top |
|
|
Harold Almonte
Joined: 04 Aug 2006
Posts: 616
|
| Posted: Mon Dec 17, 2007 7:10 pm Post subject: |
|
|
| We were so quiet with the scoring rating, then came Mike with his decision making adjust, and now this...uaao. |
|
| Back to top |
|
|
Dan Rosenbaum
Joined: 03 Jan 2005
Posts: 541
Location: Greensboro, North Carolina
|
| Posted: Tue Dec 18, 2007 12:20 am Post subject: |
|
|
| gabefarkas wrote: | | You keep mentioning this paper with Lewin, but either I missed something or it just kind of started coming out of nowhere. Is there an introductory post about it that I missed? Did you share a draft or something? Can you talk about it at all? Is there a draft abstract? Or is this the Pot/Kettle paper that is printed and in my bag and that I've been meaning to read for about a week now? |
viewtopic.php?t=1589 |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2005 phpBB Group
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
