Page 1 of 2

2017 MVP tracker

Posted: Fri Dec 09, 2016 5:19 pm
by Mike G
http://www.basketball-reference.com/friv/mvp.cgi

The above page gives a daily update on MVP Award probability based on factors leading to previous winners.

Code: Select all

prob.   MVP       tm   W%   VORP   WS
.323   Durant    GSW  .870   2.1   5.0
.303   Harden    Hou  .682   2.4   4.6
.147   Westbrook OKC  .636   3.2   3.6
.065   Curry     GSW  .870   1.7   4.0
.060   LeBron    Cle  .750   1.6   3.1

.039   Paul      LAC  .696   2.4   4.6
.020   Kawhi     SAS  .783   1.7   3.9
.014   Butler    Chi  .545   1.8   4.6
.011   Lowry     Tor  .682   1.6   3.4
.010   Love      Cle  .750    .8   3.0
Anthony Davis leads the league in Pts/G, Blk/G, and PER; but he plays for a non-contender.

Re: 2017 MVP tracker

Posted: Sun Dec 11, 2016 12:17 pm
by Mike G
Sudden switcheroo.

Code: Select all

prob.   MVP       tm   W%   VORP   WS   eWin
.333   Harden    Hou  .708   2.6   4.9   4.8
.227   Durant    GSW  .833   2.2   5.1   4.7
.142   LeBron    Cle  .773   2.0   3.8   3.8
.103   Westbrook OKC  .609   3.3   3.6   5.7
.065   Paul      LAC  .708   2.6   5.0   4.0

.045   Curry     GSW  .833   1.7   4.0   3.7
.028   Kawhi     SAS  .792   1.8   4.2   3.9
.022   Butler    Chi  .565   1.9   4.9   3.8
.018   Lowry     Tor  .696   1.9   3.8   3.4
.017   Love      Cle  .773    .9   3.5   3.2
eWins do not include last night's games.

Code: Select all

GmSc   mvp?    Pts  Rb  As
40.0  LeBron    44   9  10
29.5  Butler    31   7   5
28.9  Paul      20   5  20
27.0  Kawhi     30   6   3
21.9  Harden    18   9  16
17.9  Love      22   6   1
17.2  Durant    21   7   1
_8.9  Curry     17   8   3
LeBron's outburst last night is his first GmSc > 40 (in RS) since his return to Cle. It's #19 in his career; Harden has 8, Dirk 7, Steph and Paul 6, ..

Re: 2017 MVP tracker

Posted: Fri Dec 16, 2016 4:22 pm
by Mike G
Harden pulls away, and DeRozan bumps Butler.

Code: Select all

prob.  mvp       tm   W%     PER   WS   VORP
.457  Harden    Hou  .731   27.8   5.4   2.8
.175  Durant    GSW  .852   27.5   5.3   2.2
.111  LeBron    Cle  .750   26.3   3.9   2.2
.074  Paul      LAC  .731   27.8   5.3   2.7
.050  Curry     GSW  .852   24.8   4.4   1.9

.044  Westbrook OKC  .577   28.5   3.6   3.5
.035  Kawhi     SAS  .808   27.2   4.6   2.0
.022  Lowry     Tor  .720   22.7   4.2   2.0
.020  Love      Cle  .750   25.4   3.8   1.1
.013  DeRozan   Tor  .720   25.7   3.7    .8

----  Butler    Chi  .520   26.8   5.1   2.0
Not among the b-r.com top 20: Lowry's PER, Love and DeRozan's vorp.
Just 2 teams represent the East, with 2 players each.
Anthony Davis still leads the league in PER, Pts/G, and Blk/G. He's 3rd in DReb and 7th in DRtg. For the .333 NOP.

Re: 2017 MVP tracker

Posted: Mon Dec 19, 2016 1:06 am
by shadow
I can see the logic in Harden being number one over Durant when you consider who their teammates are, but I don't quite understand why Harden is such a commanding favorite over Durant given that Durant has better numbers across the board except for assists and FTs. Perhaps it's also factoring in the lower expectations for Houston going into this season? (which they're on pace to drastically exceed at this point)

Re: 2017 MVP tracker

Posted: Mon Dec 19, 2016 1:36 am
by Nathan
Houston's on pace to finish with 58 wins according to 538, 14.5 wins better than their preseason vegas line. Second place in that category, as of right now, is Toronto (projected 6.5 games better than their vegas line). That along with his statistical resume and head-to-head wins against GS and OKC recently makes Harden MVP so far in my eyes.

Re: 2017 MVP tracker

Posted: Mon Dec 19, 2016 3:00 am
by shadow
Right. I have no qualms with him being the favorite or eventually winning it when you factor in those details. I'm more interested in what it is in the B-R formula that puts him at a 45% to 17% advantage over Durant, considering the stats they reference on that page show a pretty clear edge for Durant in all but 2 categories. My guess is their model is penalizing Durant's chances a bit for having another MVP caliber player in Curry on his team since voters tend to look more favorably on guys who are the sole star on their team (Rose over LBJ in 2010-11 for instance).

Re: 2017 MVP tracker

Posted: Tue Dec 27, 2016 10:59 am
by Mike G
It's a bit more like a 'race'.

Code: Select all

prob    mvp     tm   Win%   PER    WS   VORP   eW82
.377  Harden    Hou  .719   27.4   6.5   3.4   17.2
.213  Durant    GSW  .844   28.2   6.7   2.8   16.4
.128  Westbrook OKC  .613   30.7   5.2   4.5   21.0
.120  LeBron    Cle  .767   26.6   4.8   2.8   15.6
.037  Curry     GSW  .844   23.9   5.2   2.3   12.6

.035  Kawhi     SAS  .806   27.3   5.6   2.4   14.1
.031  Paul      LAC  .667   27.1   5.7   2.8   13.5
.030  Lowry     Tor  .733   23.5   5.3   2.5   12.5
.017  Love      Cle  .767   25.1   4.3   1.2   12.1
.010  Giannis   Mil  .483   28.2   5.4   3.2   15.7

.     A Davis   NOP  .364   28.8   4.9   2.0   16.4
.     Cousins   Sac  .452   27.7   4.0   2.1   16.7
A couple non-contenders with nice numbers are thrown in.
Harden peeked above 50% for a day or so, then dropped to 35%. His team is now over .700, so maybe that helps.
Giannis is new to the list, bumping DeRozan. He's tied for 3rd in PER, 5th in WS/48, 3rd in VORP -- ahead of everyone in at least one of these.

The media seem to love to make it about a West star vs an East star. LeBron has taken up the Jordan hobby of crushing the MVP in the playoffs. Otherwise the west is dominant in superstars.
On Sunday, Chauncey Billups was certain this is Westbrook's party.

Re: 2017 MVP tracker

Posted: Sat Dec 31, 2016 2:38 pm
by Mike G
... I don't quite understand why Harden is such a commanding favorite over Durant given that Durant has better numbers across the board except for assists and FTs...
I think mvp voting is historically about 2 things: Team success and per-game averages. Presume that's why these elements are shown at the b-r.com tracker page.
Harden now with roughly the same chance as everyone else combined.

Code: Select all

prob  per game    W%    Pts   Reb   Ast   Stl  Blk   TO
.483   Harden   .735   27.8   7.9  11.8   1.4   .3   5.6
.197   Durant   .853   25.9   8.9   4.8   1.1  1.6   2.3
.100   LeBron   .774   25.4   7.9   8.6   1.4   .6   3.9
.077  Westbrook .606   31.3  10.5  10.6   1.3   .3   5.5
.038   Leonard  .818   24.4   6.0   3.1   2.0   .5   2.0

.036   Curry    .853   23.9   4.2   5.9   1.7   .1   2.7
.020   Paul     .629   17.7   5.3   9.5   2.2   .2   2.3
.019   Lowry    .688   22.2   4.8   7.3   1.5   .3   2.9
.018   Love     .774   22.0  10.9   1.7   1.1   .5   2.0
.009   Giannis  .484   23.4   8.9   5.9   2.0  1.8   3.0
Stl, Blk, and TO might be tiebreaker stats after the big 3. Among these, Harden leads in Ast, 2nd in Pts, middling in Reb -- which is amazing for a PG.
The Bucks are given an 88% chance of reaching the playoffs. The Anthony Davises are at 11%

Re: 2017 MVP tracker

Posted: Sun Jan 01, 2017 1:52 pm
by Mike G

Code: Select all

.564   Harden
.157   Durant
.092   LeBron
.074   Westbrook
.030   Leonard
etc
What a 53-16-17 game will do. Never mind 8 TO.

Re: 2017 MVP tracker

Posted: Sun Jan 01, 2017 8:01 pm
by Voyaging
What's interesting though is that Westbrook is still an overwhelming favorite in Vegas.

Code: Select all

Westbrook -200
Harden    +300
James     +500
Leonard   +1000
Durant    +1500
Kinda shocked by that. What do they know that we don't? It looks like basketball-reference's algorithm might just be way off.

Re: 2017 MVP tracker

Posted: Sun Jan 01, 2017 8:58 pm
by shadow
The size of the market dictates the efficiency of the lines, since the larger markets with the highest maximum bets attract the most action from professional bettors and syndicates, whose large bets drive the lines towards the true odds (once you strip out the vig) and are enough to outweigh the larger number of smaller bets made by recreational gamblers. I'm not in Vegas and I can't find MVP odds on any of the usual offshore sites at the moment, so I don't know what the maximum bet is on the MVP winner, but I'm willing to guess it's around 1K at most places. When you compare that to what you can bet offshore at Pinnacle/CRIS on NBA spreads (25K) and totals (5K), obviously it becomes more difficult for professional money to overwhelm the public money like it does in larger markets. That's also assuming professionals even bother wagering on the MVP at all. Between the lower limits and being forced to tie up money for a long time on a bet that can be derailed by a fluke injury, it's not likely a market that attracts the sharpest professionals. In this case it wouldn't be unusual for Vegas or other sportsbooks to shade the lines towards who they perceive to be the public favorite, especially if they don't consider that player to be the true favorite, in order to increase their house edge. With max bets in the 1K range their exposure to large bets from sharps is limited, especially if they don't allow re-bets from the same person/account like they typically do on spreads/totals.

Another reason these type of bets aren't that attractive to professionals is the amount of vig being charged. For a typical NBA spread/total lined at -110/-110, the house edge is 4.55%. Based on the odds you posted, the house edge for the MVP wager is 19.14% according to the calculator at the link below. Vegas will increase the vig and usually lower the max wager on a proposition when they are less certain of their predictions and also to protect themselves from gamblers who have an information edge over them (i.e. someone hacks the league office and finds out who won the MVP vote before it's announced and wants to place a large wager on that player). An MVP vote that depends on the subjective assessment of a group of reporters isn't exactly the most predictable event most years. Obviously in some years it ends up being a runaway for one player as the season progresses, but the money pouring in on that player should push the odds to their true value and the lower max bets limits the overall exposure.

House edge calculator: https://wizardofodds.com/games/sports-b ... alculator/

Re: 2017 MVP tracker

Posted: Sun Jan 01, 2017 9:09 pm
by Voyaging
shadow wrote:The size of the market dictates the efficiency of the lines, since the larger markets with the highest maximum bets attract the most action from professional bettors and syndicates, whose large bets drive the lines towards the true odds (once you strip out the vig) and are enough to outweigh the larger number of smaller bets made by recreational gamblers. I'm not in Vegas and I can't find MVP odds on any of the usual offshore sites at the moment, so I don't know what the maximum bet is on the MVP winner, but I'm willing to guess it's around 1K at most places. When you compare that to what you can bet offshore at Pinnacle/CRIS on NBA spreads (25K) and totals (5K), obviously it becomes more difficult for professional money to overwhelm the public money like it does in larger markets. That's also assuming professionals even bother wagering on the MVP at all. Between the lower limits and being forced to tie up money for a long time on a bet that can be derailed by a fluke injury, it's not likely a market that attracts the sharpest professionals. In this case it wouldn't be unusual for Vegas or other sportsbooks to shade the lines towards who they perceive to be the public favorite, especially if they don't consider that player to be the true favorite, in order to increase their house edge. With max bets in the 1K range their exposure to large bets from sharps is limited, especially if they don't allow re-bets from the same person/account like they typically do on spreads/totals.
You definitely understand better than I do, but it seems to me that the shading specifically on Harden is far too large. I've never seen that kind of shade. +300 on ~40-50% MVP odds is just absurd.

I'm much more leaning towards basketball-reference's algorithm being poor for this season than I am that Harden is the actual favorite despite being +300. But I guess we'll see how the voting (and season) goes.

Re: 2017 MVP tracker

Posted: Sun Jan 01, 2017 9:32 pm
by shadow
Well the market for this bet has been open for quite a while now (probably even before the season started). Most, if not all, of the odds movement at sportsbooks on this wager has likely been driven by money wagered by public/recreational gamblers. That betting activity is what can end up shading the odds towards the public favorite (presumably Westbrook in this case), even if he shouldn't necessarily be the true favorite. Any shading done deliberately by the sportsbooks would have occurred when betting initially opened on this prop before any wagers were placed. I don't imagine there would be a lot of odds adjustments done by the sportsbooks without the influence of money being wagered unless a player suffered a long term injury, or a player just drastically outperforms his preseason expectations early on in the year. Harden opened +1600 in July, moved up to +1000 in October right before the season started and is +300 now based on the odds you posted. So he has either received a fair amount of wagers or sportsbooks are manually adjusting his odds based on his spectacular early season performance. There's no way to know exactly why the odds are moving unless you work for the sportsbook. Just curious, which site did you find those odds at?

Re: 2017 MVP tracker

Posted: Mon Jan 02, 2017 7:53 pm
by Voyaging
shadow wrote:Well the market for this bet has been open for quite a while now (probably even before the season started). Most, if not all, of the odds movement at sportsbooks on this wager has likely been driven by money wagered by public/recreational gamblers. That betting activity is what can end up shading the odds towards the public favorite (presumably Westbrook in this case), even if he shouldn't necessarily be the true favorite. Any shading done deliberately by the sportsbooks would have occurred when betting initially opened on this prop before any wagers were placed. I don't imagine there would be a lot of odds adjustments done by the sportsbooks without the influence of money being wagered unless a player suffered a long term injury, or a player just drastically outperforms his preseason expectations early on in the year. Harden opened +1600 in July, moved up to +1000 in October right before the season started and is +300 now based on the odds you posted. So he has either received a fair amount of wagers or sportsbooks are manually adjusting his odds based on his spectacular early season performance. There's no way to know exactly why the odds are moving unless you work for the sportsbook. Just curious, which site did you find those odds at?
Sportsbook.ag

The shading is almost certainly mostly to do with his early season performance, which has b-r giving him an almost 50/50 chance to win MVP. I just can't see Vegas offering odds that are so overwhelmingly +EV for bettors.

Re: 2017 MVP tracker

Posted: Mon Jan 02, 2017 8:51 pm
by shadow
OK, if it was sportsbook.ag that makes a little more sense. They're a very shady book with a terrible reputation on SBR with a D- rating. Their review page is littered with articles about payment complaints. Their lines are usually shaded to the "square" side on most regular wagers which is typically the favorite and the over (i.e. if Warriors are -7 at Pinnacle, they'll be at least -7.5 at sportsbook.ag). They don't want any part of serious professional action. They just want squares who will bet the heavy -EV side of their shaded lines so they can take all of their money and have no intention of paying out even if you do win.

SBR review page - http://www.sportsbookreview.com/betting ... sbook.com/