Fantasy Hoops from Scratch

Editor's Note: Bwoodsxyz writes a weekly stat-oriented column for Bullets Forever.  It appears every Wednesday at noon.

In a time when everyone and their grandmother seems to play fantasy football and there seems to be as many words spent on MLB fantasy as MLB, fantasy hoops just seems to lag, even among the hoops faithful.

This issue pops up from time to time, and BDL recently ran a guest column on it that caught my attention.

There are a lot of problems with making the NBA work in fantasy.*  As the start of a thought experiment, I'm going to run through the issues involved in having an NBA fantasy league, make my suggestion for each, and see what comes out the other end.**  I know I'm going to miss some problems and come up with stupid solutions to others, so, please consider this a "first draft", and have at it in the comments, and, if it goes well, I'll produce a revised Bullets Forever fantasy hoops manifesto as the work of the community.  I know we have a good variety of perspectives contributing on this site, and maybe even some folks who have figured out how to make things work, so this should be interesting.

First off, let's touch on draft vs auction and the question of keepers.  These two areas seem largely matters of (sometimes very strong) personal preference in NFL/MLB fantasy.  I've always wondered, however, whether hoops just *should* be one way or the other in these areas, particularly keepers.  I think part of works about the NBA for the die-hard fan (and different from NFL/MLB) is following team chemistry development from year-to-year.  Starting every fantasy season from scratch is nice in that participants don't get permanently saddled with their mistakes (unlike certain NBA franchises), but it totally tosses out the idea of building a whole, piece-by-piece.  Thoughts?

Ok, let's get down to business.

League size/player pool

The problem:  There are 30 NBA teams, while most fantasy leagues have around 10-12 teams.  So, the league is deep with superstars, very deep with stars, and if a player isn't a starter or very big-minute reserve, he is basically irrelevant.  Part of the fun of being someone who falls more in the category of being a "die-hard" fan, though, is following things like players with ridiculous upside who aren't meaningful pieces yet,*** or playing time controversies on going-nowhere teams-all of the things that go into the NBA season but are below the radar of the coverage of the league as one big "who is best, Lebron, Kobe, or D-Wade" debate.  Things like that should be relevant to fantasy.

The solution:  This isn't a problem in the NFL, so we can't look there for help.  In MLB, people who care about this sort of thing have the neat solution of just breaking off by AL and NL.  It creates some issues with the realities of interleague trades, etc., but nothing insurmountable.  For about half a second, this makes breaking up by East vs West seem like a good idea.  After that, not so much.  The two halves of the NBA are just too tightly tied together for it to work out the same way.  Many baseball fans spend 90% of their attention on the league their favorite (real) team is in-that doesn't happen with basketball.  I saw this tried once for the nba, and it was a train wreck.  So, instead, I'll throw out there two suggested fixes that I haven't seen tried.

Fix 1-limit the team/player pool by general agreement. 

For example, in a 12-team league, maybe you'd want 14 NBA teams to provide a player pool.  Let's say most of the guys in the league are either Wiz or Sixers fans.  Ok.  The Wizards and Sixers are both in the pool.  To make the rest of the pool, have a "draft" of other NBA teams.  Everyone in the league gets to elect one team to add to the player pool.  That ensures that everyone can get either their "second favorite" team or the team with their favorite player in the league.  Obviously, though, there are problems with this method on a year-to-year basis if you try to combine with keepers.

Fix 2-have a huge league.  As in, 20 to 24 teams.  There are two immediately apparent problems with this, one of which I don't think is really a problem.  The easy one is that it might seem hard to round up that many people.  But, given all the online resources and die-hard gathering places, it seems to me it could be pulled off.  The other problem is that it makes for a lot of people playing but only one winner.  To this, I'd suggest breaking the league into 2 or more divisions, and adding some sort of "playoff" component using the real nba playoffs.  Large leagues could be assembled using two or more sets of friends forming their own divisions, combining to make one large league. 

Trading

The problem:  Dump trades too easily get out of hand.

The solution:  No trading after Christmas, or maybe even Thanksgiving.  Either way, that's enough time for people to realize their team has a flaw in construction that needs to be addressed, but hopefully not enough time for much of the league to start throwing in the towel. 

On the other hand, you could argue this isn't really a problem to be solved, since dump trades are very much a part of life in the real NBA. 

Roster/positions

The problem:  As Bill Simmons correctly observed in his column on fixing NBA fantasy (and this ties into the too-many-stars problem above), it is just wrong to allow a "Kevin Martin, Ray Allen, Kobe" combo as starting "guards".  There are 5 positions in basketball and there have been for a long time.  I like tweeners and creativity (Nowitzki at center, Stephen Jackson at pg) as much as anyone.  In the real NBA, that is.  In fantasy, however, this just gets out of hand way too quickly.

The solution:  I'll use the Bill Simmons solution as a starting point,

Teams have to start a point guard, shooting guard, small forward, power forward and center, with four extra rotation spots for a guard, swingman (a 2 or a 3), backup big man (a 4 or a 5) and a ninth man (everyone is eligible).

but make two changes:  first, the addition of a mandatory backup PG (I mean, come on, who would go into a game without a backup PG?), and, second, instead of the "everyone" spot, make it another backup big. 

So, the active spots are PG, backup PG, SG, backup G, SF, backup SG/SF, PF, C, and 2 backup PF/Cs.

Combined with the scope-of-league fix above, you'd mostly have real nba backups at the backup positions, which I think is a good thing.

"Streaming" of players

The problem:  Whether in roto or H-2-H, an advantage can be had by being more active in getting lots of player-games in by (1) being very attentive to making sure your players are "starting" as much as possible and (2) using and abusing the waiver wire to replace easily-replaceable players with others who will be playing more games in a coming stretch.  Too many leagues devolve into a battle of who is paying the most attention to their team and waiver wire.  Some people are fine with this, obviously, but it can be very frustrating.

The solution:  Aside from the use of % categories to take some of the usefulness away, it is also possible to cap the number of games played under either scoring method.  I guess that's ok. 

A couple other ideas:  (1) to address the "within roster" streaming issue, use only 2 more roster spots after the 10 active players, and only allow adjustments once every two weeks; (2) to address the off-roster streaming, again only allow adjustments once every two weeks, and once you cut a player you can't have that player back that season.

Scoring method

The problem:  head-to-head or roto?

The solution:  Roto.  H-2-H pretty much has to require weekly scoring.  The NBA schedule is too fluky for that.  (On the other hand, that problem should even out over a long season if you don't have people streaming players.)  My other reasons:  (1) I like roto better for MLB, H-2-H for NFL, and I think the NBA is just plain more like MLB; (2) it seems like it would be a colossal pain to try to use any advanced stats on a weekly basis, and I want to try to pull them in.

Scoring categories

The problem:  I don't even know how to begin to describe it.  I've just never been in a league where the categories made as much sense as most NFL and MLB leagues.  NFL H-2-H scoring basically just seems like it must be the way it is. 

MLB can go a lot of different ways, but generally seems to use a mix of stats tied to winning and stats that are just cool (stolen bases, I'm looking in your general direction).  The specifics of the "tied to winning" set has shifted some as the tamer of the baseball "advanced" stats have become more widely accepted, so you now see things like OBA or SLG or OPS crowd out some of the old standards like batting average, but basically it has stayed the same. 

The solution:  Win Shares and nothing else matters.  (I'm kidding.  Mostly.  A WS fantasy league might be interesting....) 

Seriously, though, go the baseball route.  That declaration is the easy part.  The hard part remains actually picking the categories.  Here's a stab at it.  Ten categories seems to work pretty well in baseball, though 12 can work too.  And, it is time to move beyond the offensive-stat bias in fantasy hoops-let's try for an even split between offensive and defensive-oriented categories.

Offense:

Points-scoring points matters, points are cool, and I think it would just be hard explaining a league where Kobe wasn't one of the best at anything.

TS%--to balance "points."  We need to reward scoring efficiency somehow anyway.  This seems like a good way to do it.

Assists-you can argue over how much they matter, but they definitely matter, and they are without a doubt cool.  I thought a little about assist % or something here, but that just seems like needless complication.

TOV--turnovers are bad.  They must be punished.  I could be talked into TOV%, but it seems like more of a pain to calculate than the benefit of the %.

Offensive Rebounds--a big man-oriented stat to balance the assists, and anyway they generate possessions so they matter.  Like with turnovers, I prefer the % on general principle, but it seems like a pain.  If there is a stat/fantasy service that would make the %s easy, I'd be tempted to go that route for each of these three categories.

(Note-I failed to carve out a niche for 3-pt shooters.  I think I'm comfortable with that.  I definitely wouldn't want to knock out any of the others to add something 3-pt related.  The more efficient 3-pt shooters will get a boost in TS% and potentially in points, so they aren't being totally dissed.  Still, a good 6th category might be 3-pt %, but with some minimum number of attempts for the team.)

I'll readily admit there's room for argument with those, but I'm happy with them on the whole.  My biggest reservation is just that guys on teams that play fast will have inflated value.  That's a big argument for going to %s across the board.  Still, that's a "problem" that everyone seems ok with in the context of players within unusually high-powered NFL/MLB offenses, so maybe I'm overstating it.

Defense:

Blocks-need I say more? 

Steals-ok, I have qualms about making Shawn Marion so valuable so quickly, but I don't know what else to do.  There are only so many viable defensive counting stats.

Defensive Rebounds-no explanation necessary. Ok, 3 down, 2 more to go.

Umm, I'm kind of stumped here.  I'll throw out some ideas, but maybe we should rethink the 5-defensive-stat thing.  Maybe Defensive Win Shares?  Maybe Defensive Rating (from basketball reference)?  Maybe Net On/Off Ct Defense (from 82 games)?  Maybe Opposing Player PER (from 82 games, looking only at the position the player is designated as)?  Any thoughts?

*  I think this understates the problems, if anything.  I'll use myself as an example.  I watch a decent number of games all season long.  I can probably name at least 10 guys on every nba roster and tell you something about their game.  I've been writing a team-centered stats column (in the off-season!).  I play in an auction AL-only roto baseball league that uses nearly every real player, even though I don't follow MLB nearly as closely as the NBA.  I *should* like and play NBA fantasy.  Instead, I have totally lost interest nearly every season, regardless of whether I happened to be in first place or last..

I know the problem isn't just me.  I've played with a bunch of different folks over the years, including many who range somewhere between "solid fan" and "crazy obsessed fan" level of NBA fandom.  Even so, I have yet to play in a league where more than a couple of people made it through the whole season.  The last few years I've been in groups that tried to have only people who are "full-season" nba fans, thinking the lack of nba attention span might be the problem..  That didn't cut it.  We've tried fixes:  tinkering the stat categories, using different systems of keepers, limiting roster moves, etc.  Nothing has made much difference.  I'm sure some of you have experienced successful runs of NBA fantasy, but, as I write this, I don't actually know anyone who has.

**  Ground rule:  this is not done with "casual" fans in mind.  I know NFL fantasy has helped "grow" the sport by bringing in more interested casual fans, but whatever.  I just can't get my mind around that working in the NBA.  And, anyway, I have a hard time believing that something could really catch on with casual fans if it couldn't first pass the smell test with folks who would read an NBA blog in that oh-so-exciting time between the end of open season on free agents and the start of training camp. 

***  Ok, I confess, this is just an elaborate way of saying that I want to be able to pick up Blatche and Young without forfeiting all chance of fielding a competitive fantasy team.