Sunday, January 20, 2013

CFB 2012 Review

I didn't have time to make any official college football picks on the blog, but here are a couple preseason predictions that were accurate and inaccurate for CFB.

My preseason prediction for the NC game was entirely inaccurate; it was Oklahoma vs USC.  Oklahoma ended up with ten wins and three losses to extremely quality opponents.  USC was extremely disappointing..  They finished 7-6.

Predictions for my favorite teams were a bit more accurate, OSU finished 12-0 which was a game within my prediction of 11-1.  They had a lot of close match-ups against inferior competition, so I certainly never thought that (even if they were eligible) they should play in the NC game.  Braxton Miller provided some great QB play.

I thought that somehow the Cyclones would get to 6-6 and another bowl game, and somehow Paul Rhoads led them to 6-6.  Other than their game against Texas, and their game against OKSU, the Cyclones either won or had a chance to win every game.  As a sidenote, I don't understand why CFB teams ever have to face teams they already beat in the regular season in bowl games, as was the case for ISU against Tulsa.

I also think that the BIG '12' is the best CFB conference top to bottom, so while some other analysts were predicting that WVU and TCU would have some major impact on the BIG 12 title race, I happened to place them both as secondary type Big '12' teams, and hence nailed both West Virginia's and TCU's final record.

Here is my top 30 rankings for CFB.  Personally I don't understand why Alabama and Notre Dame were playing for the NC.  Alabama got dominated @home against Texas A & M (though credit to Bama they did fight back) and other then @LSU beat very few very good teams.  Notre Dame beat Stanford and Oklahoma, but they also took 3 OT's to beat Pittsburgh.  Pittsburgh.  Not even a Northwestern type of school, but Pittsburgh...  Imho, it should have been Texas A&M vs Oregon in the NC.  Anyways, on to my rankings.

1. Oregon
2. Texas A&M
3. Alabama
4. Ohio State
5. Kansas State
6. Georgia
7. Notre Dame
8. LSU
9. Stanford
10. Oklahoma
11. Florida
12. Florida State
13. South Carolina
14. Clemson
15. Louisville
16. Texas
17. Boise State
18. Northwestern
19. Oregon State
20. Utah State
21. Vanderbilt
22. Nebraska
23. Michigan
24. Baylor
25. Oklahoma State
'Other's receiving votes'
26. Northern Illinois
27. San Jose State
28. Penn State
29. Tulsa
30. Cincinnati

Monday, January 7, 2013

OKC vs Miami QS vs NQS Explained, Part 2



I originally wrote this post not too long after the NBA finals, so another post right on time..

The contents of the PDF are set up a little messily, but with a little guidance it should be somewhat self explanatory.  (Column B1) 'q' simply means a quality two point shot taken, and the player's average points per attempt.  (C1) 'nq' is a non quality two point shot.  (D1) 'raw 2' means the weighted value of both NQ and Q shot attempts.

(E1) 'q3' a quality three point shot.  I'm going to guess you can figure out (F1) and (G1).  So on to (H1).  It's blank, so I guess I forgot to put something like, 'raw value'.  So (H1) is simply the weighted value of any shot attempt by a player.    (I1) 'Usage' the percentage of possessions that end with that player taking a shot.

It's a little tedious, but I guess I'll go through an entire line. So for instance; (B3) a Russell Westbrook quality two point attempt 'will' result in 1.32 pts.  (Hopefully you're following what I'm saying.)  You could also say that RW would make 66% of his quality two point attempts.  (C3)  RW scores a raw 0.71 points on any NQ 2 point shot taken, or 35.5%.  (D3) Westbrook's raw 2 point shooting is 1.05, or 52.5%.  (E3) Westbrook did not make any quality threes, so his average for a q3 is 0, and Idk if you can figure this out.. heh-heh, but that means he shot 0% on quality three point attempts.  (F3)  0.46 pts per shot for a nq3, or roughly 15.3%.  (G3)  A raw 3 PPS of 0.33, or 11%.  (H3)  And his pure value of any given shot is, 0.93, or 46.5%.  (I1)  Any random Thunder possession ended with Westbrook taking a shot approximately 32% of the time.

That should clear up everything through the first 7 lines, except for cases like (C7), and (E7).  In (C7)  a * means that Ibaka did attempt nq2's, but the attempts were minimal.  In the case of (E7, or any of the other *'s) a * means that the shots were too negligible to record on the pdf, because the sample size was way too small.
And occasionally an anomaly made it through anyways, since I was only doing these stats off of 5 games. Like in B15 you could draw the (wrong) conclusion that since Chalmers averages 2.00 points per quality two point attempt, any quality 2 that he attempts will always be made.  (which is, obviously, not true)

(The reason for the colors)  (B8-H8)  is sort of more complicated..  And to make it even more complicated...  I made some mistakes in setting that up.  Feel free to ignore the first number in each column (2, 3, 4, 4, 7, 8, 9).   The red (B8) means a spot where the shooter is hot  (as compared to the rest of the league).  The blue (H8) is a cold shooter.  (as compared to the rest of the league).  Instead of going from cold to hot, (as I kind of wish I had), I went from hot to cold.  So the far left side (B8) is a lot better than the far right side (H8)..  While something like (D8 compared to E8) would only be slightly better.  Get it?

For Lebron to have a league average pure shot value for a Q2, his shot value could be anywhere within the standard deviation of .1 from 1.30.  Since Lebron's Q2 shot value is only 1.36 and the .1 deviation covers that, this means that (based on all of 5 games) Lebron has effectively a league average pure Q2 value.

I was comparing each individual shooter to my league-wide NBA average.  (Found at row 38 and on down)  I think the colors can make it easier to gauge how well players shoot, it adds some detail to the pdf..  But it also adds some clutter to it, and it (like everything else) takes quite a while to add.
     As a quick example of the color... Looking at Lebron James, both his Q and NQ 2 pt attempts are effectively league-average, but on the plus side, his raw 2 point-value, and his Q3 are slightly above league average, but, on the downside, his raw 3 pt-value is frigid.  I'll assume that makes sense to you, and move on..

The only other note before the end of this section, is that the 'metamorphosis of Thunder/Heat' (A5 and A14), is just a representation of the rest of their players, the rest of the players who are not named, with their stats combined.

I thought this was somewhat fascinating.  If I do this again, then hopefully I can find a better way to attach the PDF.  It looks like if you click on it, then it's a bit easier to see.

OKC vs Miami Quality Shots vs Non Quality Shots, Part 1



















Saturday, January 5, 2013

NFL Playoffs Predictions 2013


1st game    win            SBA               SBW
Denver 65% 2 1
NE 65 4 1
Houston 65 2.5 2
Baltimore 90 1.5
Indy 10
Cincy 35
NFC
Atlanta 80% 5.5 4
SF 40 1
GB 90 2 1.5
Washington 45
Seattle 55 1.5 0.5
Minnesota 10