Top values on new metric called "Individual Player Values"

[talkingpractice]

We put new values up last night, and I'm generally pleased with them.

Re: my comment that one number metrics aren't that important, I think I meant that much less formally than it came across. What I meant to say basically is that basketball is a game between 2 teams of 12ish guys that get split up into various 5 player lineups, to play against opposing 5 player lineups, with some constraints subject to how many minutes the individual players can play within a game (and consecutively). Shrinking all that down to the player level is fun in the sense that it's nice to discuss who the best individual players are. But who the best individual players are isn't really super relevant (at least in terms of wagering on the teams). We care more about the lineups that will be out there that night, how long those lineups will be out there, how those lineups play vs the lineups of the other team, and how any individual matchups look. The individual player values in and of themselves don't really come into play that much. But I do agree that it's definitely a lot more fun to talk about whether or not RWB is now better than DWade, who is better between {Chandler, M Gasol, Howard}, etc.

re: some of Crow's comments, we have Deron Williams at -2.6 on defense too, and he's always been -1.5 or worse for us. I speculate that we like DWill more than xRAPM does because we're using a pretty sticky prior for his offensive ability. We have Mayo > +1 now, and I'm happy with that.

When I have a minute soon, I'll probably make a list of the players we're in most disagreement with compared to xRAPM or ASPM, to see if there's anything interesting we can learn from that.

[Crow]

Yes, comparison of different metrics for players can be interesting and helpful.

A master chart for top or all players on the older and newer metrics would be nice. If anyone assembles that, that would be a nice thing to share.

[talkingpractice]

We've now greatly expanded the amount of information on the site, and tried to include the things for which we received the most email requests. Things up there now

1, List of all players by IPV (predictive)
2, List of all players by total value produced this year (descriptive, ie basically an MVP ranking)
3, Top players by position
4, Top 10 rookies by IPV and by value produced this year

Same caveats as before apply.

People have asked for comparisons of IPV values to values from other metrics. The only other all-in-one metrics that I'd personally look at for more than 30 seconds would be xRAPM or ASPM, and I of course know roughly where the commonalities/differences lie between each of those and IPV. Differences with xRAPM are mostly regarding our liking some of the high usage/offense guys more than xRAPM does, so we have people like Ray Allen, Kyrie, Lillard, Kobe, and similar as rated way more highly than xRAPM does. On the other end, we have some of the big defense guys lower than xRAPM does, and we have Manu much lower. If anyone wants us to provide the full sample of differences from IPV to xRAPM and/or ASPM for the whole league, then we can probably do that later this week.

[talkingpractice]

All of our data was updated on the 27th, and we've also included a Playoff Power Ratings sheet. Sorry that we've been slow doing this, we'll begin updating them twice/week again starting on Monday. We've had some requests for daily updates, and I guess that's something that we can setup in the near future (maybe over the ASB), but as of now this is the best I can do.

[talkingpractice]

I'm sorry to everyone who asked... we've been a bit lazy getting the numbers updated, and will do it within a couple of days (and will have a much better setup going fw so that it will be updated every few days from then on).

In the interim, we did get a girl to get rid of the Vulcan flare of the site and make it look human.

[Crow]

Thanks again for the extra effort to share the new metric information.

[talkingpractice]

Happy to provide. I've learned so much from this forum over the years.

[talkingpractice]

Updated as of Jan 14th, and I'm sorry for the delay. From now on, someone will update it every few days.

About 9234 people have asked me for themes/teams we're trading alot now , and that's something that of course I'm not going to discuss very often. However since this market is so tiny and we did all the dirty work already... we did just finish buying up the Bulls to win the title about wherever we could. They match up so well with the Knicks (Deng is Melo Kryptonite, Noah can stay with Chandler, there is a huge gap in coaching, and lol at Felton or Kidd/Prigioni trying to stay with Rose), and some places had them as high as 40:1 for awhile (they're still at around 20:1 in a lot of places now). I surely don't expect them to actually win the title, but the bet is really more like a bet on them to have a good chance to make the conference finals (the futures move and are liquid in real time).

[Crow]

I can see that the Bulls would be considered attractively priced to date. Looking just briefly, I didn't see any other teams that looked like strong buys for the odds currently given to me. I see the Clippers at 6-1 one place. They were probably a good buy some weeks ago when the odds were longer.

[talkingpractice]

Updated through last night. There are 3 guys who are reallyyyyy separated from the rest of the multiverse right now.

Lebron +8.75
Paul +7.50
KD +7.44
.
.
.
.
.
then big dropoff to Duncan, Parker, and Wade (all approx +5).

[Statman]

Updated through last night. There are 3 guys who are reallyyyyy separated from the rest of the multiverse right now.

Lebron +8.75
Paul +7.50
KD +7.44
.
.
.
.
.
then big dropoff to Duncan, Parker, and Wade (all approx +5).

I think almost every metric here will have those as the top 3. Mine does - at least the quick crunching I did a few days ago after reworking (streamlining for faster future 2013 updates & historical season updates) my NBA spreadsheets.

I really like Crow's idea - my NBA player ratings/rankings will be posted (debuted) on my site at the latest Monday - I have zero problems with the results being listed/compared to others. There have to be a good 10 different player metrics floating around this board?

[v-zero]

I think almost every metric here will have those as the top 3. Mine does - at least the quick crunching I did a few days ago after reworking (streamlining for faster future 2013 updates & historical season updates) my NBA spreadsheets.

Mine based on an advanced box score (unfinished as it is) has a thing for versatile defensive big men, and has a tendency to penalise a lot of blocking that forwards do because it appears that blocks that aren't close to the bucket indicate bad defensive ability - possibly because these guys are the guys giving away charity shots on long jumpers that were low percentage shots in the first place. Whatever the case, the numbers point towards players like Duncan and Noah, defensively, being quite a cut above Durant and LeBron. But yes, Chris Paul does show up as the best offensive weapon in the league.

[talkingpractice]

I really like Crow's idea - my NBA player ratings/rankings will be posted (debuted) on my site at the latest Monday - I have zero problems with the results being listed/compared to others. There have to be a good 10 different player metrics floating around this board?

Yeah, I'm happy to provide our numbers for anything like this, too. We update the site every few days now, and we're fine trying to accommodate whatever format/system someone were to create, if anyone wants to compile rankings from various metrics.

[deepak]

Are you computing offensive and defensive ratings separately, and then adding together for the final IPV rating? If so, could you also post the offensive and defensive ratings?

[EvanZ]

ezPM is always available at my site, of course:

http://www.d3coder.com/thecity/ezpm-player-ratings/