Saturday, April 28, 2012

Today's Games - going dark temporarily

The Today's Games feature for the NBA will be going dark temporarily.  The default methodology I use doesn't work that well for the playoffs.  However, I hope to launch a playoff version of the Ticker soon (with daily updates), so check back.  I also have MLB versions of the rankings in the works as well that I hope to launch soon (with corresponding Ticker and Today's Games features).

Saturday, April 21, 2012

NBA Team Rankings - April 21, 2012


Tank primeaguerra
World War I Tank
Here are the NBA Betting Market Team Rankings as of April 21, 2012. These rankings use the Vegas point spreads from recent games to derive an implied measure of team strength (Generic Points Favored, or GPF).  By combining the point spread information with the betting over/unders, I can also decompose team strength into its offensive and defensive components (oGPF and dGPF). For more detail, see the Methodology page.


At this point in the season, it's hard to say exactly what these rankings are telling us. Are the Thunder and Spurs really that much better than the Heat and Bulls?  How much are the point spreads affected by teams resting up for the playoffs?  Are teams that are out of the playoffs "tanking" in the hopes of getting a better lottery pick in the draft?  Or do they simply not try as hard at this point?

Saturday, April 14, 2012

NBA Team Rankings - April 14, 2012

Derrick Rose 2
Derrick Rose
Here are the NBA Betting Market Team Rankings as of April 14, 2012. These rankings use the Vegas point spreads from recent games to derive an implied measure of team strength (Generic Points Favored, or GPF).  By combining the point spread information with the betting over/unders, I can also decompose team strength into its offensive and defensive components (oGPF and dGPF). For more detail, see the Methodology page.


The Thunder retain the top spot this week.  Both the Celtics and the Hawks moved up in the rankings significantly.  The Hawks, in particular, are an interesting case.  Despite having a top 10 Win/Loss record for most of the season, they were consistently ranked in the 15-20 range according to the point spreads.  And then they jump 3 points in GPF in the course of a week (see for yourself on The Ticker).  It's as if the market suddenly snapped to attention.

Thursday, April 12, 2012

The Ticker - Updated Features

I have restructured the Ticker feature of this blog to (hopefully) make the data easier to view and navigate.

  • I have separated the various charts into their own pages
  • I added a GPF-only graph because I felt that the old format was masking the GPF movement with the additional oGPF and dGPF lines (this old format is still accessible under the "GPF+oGPF+dGPF" link)
  • I added a Generic Win Probability (GWP) graph, which just converts the GPF graph into a win probability scale
  • The Generic Over/Under (GOU) graph has been split out into its own page
  • There is now a link to the rankings in table form, which is updated daily.  It has all the metrics I provide in my weekly blog post of the rankings
  • There are placeholders for other sports and prior seasons.  I hope to have those live soon

new version of the Ticker: link
background: original launch

Saturday, April 7, 2012

NBA Impact of Rest - Update

I recently discovered that my analysis on the impact of rest on NBA point spreads had an error that was skewing the results.  The original analysis found that a team that was playing on no rest was penalized by 0.75 points in the point spread.  After the correction, the penalty changed from 0.75 points to 1.25 points.


The new 1.25 point penalty was factored into the Today's Games and Ticker features earlier this week.

NBA Team Rankings - April 7, 2012


Oklahoma City 1890
Here are the NBA Betting Market Team Rankings as of April 7, 2012. These rankings use the Vegas point spreads from recent games to derive an implied measure of team strength (Generic Points Favored, or GPF).  By combining the point spread information with the betting over/unders, I can also decompose team strength into its offensive and defensive components (oGPF and dGPF). For more detail, see the Methodology page.


The Thunder have displaced the Heat from the top of the rankings this week, despite having lost three straight (including one to the Heat on Wednesday). The Thunder's last five games have been against the Pacers, Heat, Grizzlies, Bulls, and Lakers.  That they were only able win two out of those five shouldn't reflect too poorly on them given the strength of the competition they were facing.

Brackets: Post-mortem


Prior to the NCAA tournament starting, I published four different brackets based on my betting market rankings.  Here's how they fared:


The first bracket I published was part of the March 12 rankings post.  This bracket was based on mechanically picking the higher ranked team for each matchup.  I then published three more brackets in this post, with an eye towards picking some "optimal" upsets that could help separate you from the pack (with a little luck).


I entered all four of these brackets into Yahoo's Tourney Pick'em challenge, and am using Yahoo's 1-2-4-8-16-32 scoring system to evaluate them.


Risk LevelBracket NameYahooPercentileCorrectChampion
BoringBetting Market Analyticslink95th45 out of 63Kentucky
RiskyWeekend in Vegaslink55th38 out of 63Ohio State
RiskierBet the Farmlink30th30 out of 63Ohio State
RiskiestJerome Kerveillink35th27 out of 63Ohio State


My self-described "boring" bracket actually performed pretty well, beating 95% of the Yahoo brackets.  This is largely due to picking Ohio State and Kansas to upset #1 seeds Syracuse and North Carolina to make it to the Final Four (and picking Kentucky as champion, of course).  But those were hardly unique picks, as you would have picked the same if you had followed the rankings according to Ken Pomeroy or Team Rankings' "Best" Bracket.

The "riskier" brackets did not fare as well, although I was true to my word when I said: "Getting risky will give you a small probability of winning the pool, and a large probability of failing spectacularly".

Sunday, April 1, 2012

NBA Team Rankings - March 31, 2012

Spurs cowboy crockett
They jingle jangle
Here are the NBA Team Rankings as of March 31, 2012. These rankings use the Vegas point spreads from recent games to derive an implied measure of team strength (Generic Points Favored, or GPF).  By combining the point spread information with the betting over/unders, I can also decompose team strength into its offensive and defensive components (oGPF and dGPF). For more detail, see the Methodology page.

For the first time this season, the Bulls have fallen out of the top 3, having been replaced by the Spurs.  I'm somewhat surprised that the Bulls haven't fallen further, given they've been without Derrick Rose for 10 straight games now.

The Knicks' ups and downs continue.  Having reached the top 5 last week, they've since fallen to #10, and will most likely fall further with both Amare Stoudamire and Jeremy Lin probably out for the season.  Follow the Knicks' progress on the Ticker.